What should a plant do when condition monitoring signals disagree?
A vibration alert, a normal bearing temperature and a clean oil sample do not necessarily contradict one another. Each technique detects different physical conditions and has different limitations. The right response is to verify the evidence and the operating context before declaring the asset healthy or failed.
Check whether the measurements describe the same state
Record speed, load, product, process temperature, flow, machine configuration and time of each reading. Compare steady state with steady state rather than startup vibration with a later stabilized oil sample. Confirm that each point belongs to the correct asset and bearing position. A tag mapping error can make two valid measurements appear contradictory.
Test measurement quality before interpreting failure
Inspect sensor mounting, orientation, calibration status, sampling location and data acquisition settings. A loose vibration sensor can create a false trend; a temperature sensor may be too far from a developing localized defect; a contaminated oil sample can create an unrelated alarm. Do not discard an outlier solely because other channels are normal. Repeat or independently verify the suspect measurement using an approved procedure.
Relate each signal to a plausible failure mode
- Vibration: may reveal imbalance, looseness, misalignment or bearing-related patterns when measured and interpreted correctly.
- Oil analysis: may reveal contamination, lubricant condition or wear debris, depending on test and sampling method.
- Temperature: may show thermal stress but may respond late or be influenced by ambient and process conditions.
- Process data: can reveal changing flow, pressure, power or efficiency that changes the meaning of a mechanical signal.
None of these readings alone identifies every failure mode. Use machine design, operating history and inspection evidence to develop and test hypotheses.
Example: a pump with rising vibration
A pump shows a rising vibration trend at one bearing, while oil analysis and bearing temperature remain within their usual ranges. The team verifies transducer mounting and compares measurements at similar flow and speed. It then checks spectrum features, suction conditions, coupling condition and recent maintenance. The result may justify closer monitoring, a planned inspection or immediate action depending on severity and consequence. It does not justify ignoring the vibration solely because oil and temperature are normal.
Define a decision and reassessment point
Record the current hypothesis, confidence, missing evidence, operating restrictions if any, accountable reviewer and next measurement time. Use asset criticality to set urgency. If a protective trip or operating limit is involved, follow site procedures rather than using a diagnostic interpretation to bypass protection.
Learn from the outcome
After inspection or repair, compare the actual defect with each signal. Update alert thresholds, sensor locations and inspection routes where evidence supports a change. A useful monitoring program learns which combinations of signals lead to actionable defects and which create avoidable alarms.
Explore KANE consulting
Discuss practical maintenance and reliability improvements for your facility.
Request a consultation →Related answers: condition monitoring technique selection · condition monitoring baseline
Technical references
DOE Operations and Maintenance Best Practices · DOE Federal Energy Management Program