Understanding the FMEA risk priority number calculator — concept and principle
Failure Mode and Effects Analysis (FMEA) is a structured way for a team to list how a product or process can fail, what the effect would be and how well current controls catch it. Each failure mode is rated from 1 to 10 for severity, occurrence and detection, and the risk priority number (RPN) combines the three into one score used to rank which failure modes to work on first.
How it is calculated
RPN = S × O × D
- Severity (S): how serious the effect is for the customer or user, 10 being hazardous without warning.
- Occurrence (O): how likely the cause is to happen, 10 being almost certain.
- Detection (D): how unlikely current controls are to detect the problem before it reaches the customer, 10 being no chance of detection.
The result ranges from 1 to 1,000. The calculator flags severity 9 or 10 for action regardless of RPN, and labels RPN ≥ 100 as high priority.
The principle
Risk in its simplest form is consequence times likelihood. FMEA adds a third factor, detection, because a defect that is always caught in-house costs far less than one that escapes. Multiplying the ratings makes any single very bad factor raise the score. The ratings are ordinal, however: a 6 is not "twice as bad" as a 3, so the RPN is a ranking aid, not a measured risk. That is why the AIAG and VDA FMEA Handbook (2019) replaced RPN thresholds with Action Priority (AP) tables, which weigh severity first, then occurrence, then detection.
Worked example
- Inputs: S = 8, O = 4, D = 5.
- RPN = 8 × 4 × 5 = 160.
- Severity is below 9, so the severity rule does not trigger.
- 160 ≥ 100, so the note reads high — prioritise.
| Severity (1–10) | RPN |
|---|---|
| 2 | 40 |
| 4 | 80 |
| 6 | 120 |
| 8 | 160 |
| 10 | 200 |
| 12 | 240 |
| 16 | 320 |
How to read the result
Use the RPN to sort failure modes within one FMEA, not to compare different teams or products whose rating scales differ. After an action, re-rate the item: a better test lowers D, a design or process change lowers O, and usually only a design change lowers S. The 100 cut-off is a common in-house convention, not a universal standard. Always review high-severity items first, even when their RPN is small; S = 10, O = 1, D = 1 gives RPN 10 but can still mean a safety hazard.
Common mistakes
- Setting a single RPN threshold and ignoring everything below it.
- Lowering the RPN by improving detection only, when the cause could be prevented.
- Forgetting that different combinations give the same RPN: 8 × 4 × 5 and 10 × 2 × 8 are both 160.
Frequently asked questions
What RPN is considered high?
There is no official value. Many companies use 100 or 120 internally, but current guidance recommends Action Priority instead of a fixed threshold.
How is Action Priority different from RPN?
AP looks up the S, O and D combination in a table that gives High, Medium or Low, with severity carrying the most weight, so it avoids the equal-weight problem of the product.
Sources: AIAG & VDA, FMEA Handbook, 1st edition (2019) · IEC 60812:2018, Failure modes and effects analysis (FMEA and FMECA)