Preventive and Predictive Maintenance
What this domain tests
This domain tests task selection and timing: preventive maintenance, condition-based maintenance, predictive technologies, failure finding, and PM optimization. The exam distinguishes maintenance that prevents failure from maintenance that merely consumes resources.
Core rules and concepts
- PM is justified when a failure mode has a measurable deterioration pattern and the task can detect or prevent it in time.
- Condition-based maintenance uses vibration, thermography, ultrasound, oil analysis, and performance monitoring to act on condition.
- Predictive maintenance forecasts remaining useful life; condition monitoring detects the onset of a defect.
- PM optimization challenges each task: Does the failure mode exist? Is the task applicable? Is the interval right? Does the task cause failure?
- Failure-finding tasks protect hidden functions, such as emergency shutdowns, where failure is not evident until demanded.
- Run-to-failure is valid for low-consequence, non-critical items where prevention costs more than failure.
Worked example
A quarterly PM requires greasing 60 motor bearings. Records show several bearing failures shortly after greasing, and vibration data show over-lubrication. The reasoning: the task is not failure-mode-based. A better tactic is a quarterly ultrasound inspection with grease only when the bearing indicates need, plus a correct grease quantity and method. The PM interval should then be based on observed condition and failure history. This removes a task that induces failure while preserving protection.
Traps that cost marks
Assuming more PM always improves reliability. Using calendar intervals with no failure-mode basis. Applying vibration monitoring to a failure mode it cannot detect. Setting an inspection interval longer than the P-F interval. Ignoring hidden failures that require failure finding. Keeping a PM because “we have always done it.”
Exam tip
For any PM question, identify the failure mode and consequence first. If the task cannot detect or prevent that mode, the correct answer is usually to redesign the task, change the tactic, or run to failure.