FMECA: what it is and how to apply it

FMECA is a structured method for determining how a system can fail, what the consequences are and which failure modes deserve the most attention. The abbreviation stands for Failure Mode, Effects and Criticality Analysis. For technical organisations, FMECA is one of the fundamental tools for setting up maintenance and reliability on a sound basis rather than on gut feeling.

This article explains what FMEA and FMECA are, why you carry out the analysis, how to build it step by step and how to translate the results into a concrete maintenance plan.

FMEA and FMECA: what the C adds

FMEA (Failure Mode and Effects Analysis) identifies, per component or function, which failure modes can occur and what effects they have on the system and its surroundings. The result is a structured overview of "what can go wrong and what happens if it does".

FMECA adds an explicit criticality analysis: the "C" for Criticality. Instead of treating all failure modes as equal, you quantify how critical each failure mode is, through a combination of severity and probability, optionally supplemented with detectability. This produces a ranking that shows which failure modes actually dominate your risk. FMEA tells you what can happen; FMECA also tells you what matters.

Why you carry out a FMECA

A well-executed FMECA serves several purposes at once:

• Determining criticality: you objectively reveal which components and failure modes have the greatest impact on safety, availability, quality or cost.
• Underpinning the maintenance strategy: for each failure mode you decide specifically whether preventive, condition-based or corrective maintenance is appropriate, instead of applying the same approach everywhere.
• Prioritising risks: with limited time, budget and manpower, you focus attention where the return is highest.

FMECA is therefore not an end in itself but a means to allocate scarce resources sensibly. See also the broader context of reliability and FMECA.

The steps of a FMECA

1. Functions and system boundaries

Start by defining the system boundaries and naming the functions. Describe what an installation, subsystem or component must do (for example "pump medium at a flow rate of X"). A clear function definition is essential, because a failure mode is by definition the loss of a function. Deliberately choose an analysis level: too high and you miss relevant failure modes, too low and the analysis becomes unworkable.

2. Failure modes

For each function, determine how it can fail. A failure mode describes the way in which a function is no longer fulfilled: a bearing seizes, a seal leaks, a sensor drifts, a valve fails to close fully. Be specific and distinguish failure modes that have different causes or consequences.

3. Effects

For each failure mode, describe the consequences at several levels: local (on the component), at system level, and on the surroundings or the business process. Think of production downtime, safety risks, quality loss or consequential damage. The effects will later determine the severity score.

4. Causes

For each failure mode, trace the underlying causes: wear, contamination, incorrect loading, assembly errors, material fatigue. The cause determines whether, and how, you can prevent or detect a failure mode. Without insight into causes, the choice of measures remains guesswork.

5. Assessment: severity, occurrence, detection and RPN

Now you score each failure mode. Three dimensions are common, usually on a scale from 1 to 10:

• Severity: how serious are the consequences if the failure mode occurs?
• Occurrence: how often is the cause expected to occur?
• Detection: how well can you notice the failure mode, or its onset, in time? Note: a high score here means poorly detectable.

The product of these three gives the Risk Priority Number (RPN): RPN = severity × occurrence × detection. The RPN therefore ranges from 1 to 1000 and offers a relative ranking. In a strict FMECA you often use a criticality score based mainly on severity and probability; the RPN is the most widely used practical variant. Treat the outcome as an aid to prioritisation, not as absolute truth: a failure mode with extremely high severity deserves attention even if its RPN is moderate.

6. Measures

For the failure modes with the highest criticality you define measures. These can address severity (change the design, limit consequences), occurrence (remove the cause, intensify maintenance) or detection (inspections, condition monitoring, sensors). Assign an owner and a deadline to each measure, and reassess the score once the measure has been implemented.

From FMECA to a concrete maintenance plan

The real value of a FMECA only emerges when it is translated into maintenance. For each critical failure mode you determine the most suitable maintenance type:

• If the failure mode is predictable and time-related, preventive maintenance (fixed intervals) is the obvious choice.
• If the failure mode shows a measurable onset, condition-based maintenance is more effective: you intervene based on vibration, temperature, oil analysis or other indicators.
• If the consequences are limited and the cost of prevention is high, deliberate corrective maintenance (run to failure) may be the wisest choice.

This links each measure to a task, interval or measurement limit. See also the difference between preventive vs corrective maintenance. The resulting tasks are then often managed in a maintenance management system; read more about what a CMMS is.

Practical tips and pitfalls

• Too detailed: anyone who analyses every screw drowns in work and loses overview. Choose an analysis level that matches the decision you want to support.
• Subjective scores: severity, occurrence and detection are often filled in by feel. Use clear scoring criteria and, where possible, failure data to substantiate the scores.
• A one-off exercise: a FMECA that disappears into a drawer after commissioning quickly becomes outdated. Treat it as a living document that you update after incidents, changes and new insights.
• Analysis without follow-up: scores without measures achieve nothing. The criticality ranking is a starting point, not an end product.

Relationship with RCM

FMECA often forms the analytical heart of RCM (Reliability Centered Maintenance). RCM uses the failure modes, effects and criticality from the FMECA and adds a decision tree to select the most effective maintenance task for each failure mode. You can apply FMECA on its own, but within an RCM programme the analysis gains an explicit link to maintenance decisions. FMECA provides the insight, RCM the systematic approach to turn it into a maintenance concept.