How RPN works
Each failure mode gets three ratings from 1 to 10. Multiplied together they give a Risk Priority Number from 1 to 1,000. Higher RPN means higher priority — but RPN has a well-known weakness, covered below.
Rating guide
Rating scales should be agreed by your team before scoring starts. The guide below is a general starting point, not a published scale — adapt the wording to your equipment and consequences.
| Rating | Severity (S) — how bad is the effect? | Occurrence (O) — how likely is the cause? | Detection (D) — will you catch it before it fails? |
|---|---|---|---|
| 9–10 | Safety or environmental hazard; regulatory breach | Happens very often; near-certain over the period | No way to detect before failure |
| 7–8 | Major production loss or major damage | Happens frequently | Unlikely to detect; checks are infrequent or indirect |
| 4–6 | Moderate loss; repair needed, some production impact | Happens occasionally | May detect with routine inspection or PM |
| 2–3 | Minor; little or no production impact | Rare | Likely to detect — monitored or alarmed |
| 1 | No noticeable effect | Remote; essentially never | Almost certain to detect before it matters |
Note the direction of Detection: a high D means the failure is hard to catch. It's the most common scoring mistake.
Why RPN alone can mislead
RPN treats the three ratings as equally important, and they aren't. A failure mode with S=10, O=2, D=2 scores an RPN of only 40, while a nuisance with S=4, O=5, D=5 scores 100. Ranked purely by RPN, the safety hazard sits below the nuisance. That's why this tool flags every failure mode with severity 9 or 10 regardless of its RPN.
The 2019 AIAG & VDA FMEA Handbook moved the automotive industry away from RPN thresholds to an "Action Priority" approach that weights severity first, then occurrence, then detection. If your organization follows that handbook, use its Action Priority tables for formal FMEAs; this worksheet's RPN and severity flag are a practical first pass.
Running a useful FMEA
- Score as a team. Include someone who operates the equipment and someone who repairs it. Their ratings rarely match at first, and the discussion is where the value is.
- One cause per row. If a failure mode has two causes, give each its own row — they usually need different actions.
- Use your own failure history. Work order and failure data from your CMMS should drive occurrence ratings, not memory.
- Re-score after actions. An FMEA that's never updated is a document, not a tool.
FAQ
What RPN is "too high"?
There's no universal cutoff, and fixed thresholds are one of the main criticisms of RPN. Rank the list, act on the top items and every high-severity item, and focus on reducing the biggest numbers.
What's the difference between FMEA and FMECA?
FMECA adds a criticality analysis on top of FMEA, ranking failure modes by both severity and probability. It comes from military and aerospace practice (MIL-STD-1629A). For most plant maintenance work, an FMEA with a severity flag covers the same ground.
Where is my worksheet saved?
Only in this browser on this device. Nothing is sent to a server. Export to CSV to keep a copy or share it.
Sources & standards
- Automotive Industry Action Group (AIAG) & Verband der Automobilindustrie (VDA). FMEA Handbook, 1st ed., 2019. — Introduced Action Priority in place of RPN thresholds.
- SAE International. J1739: Potential Failure Mode and Effects Analysis (FMEA).
- U.S. Department of Defense. MIL-STD-1629A: Procedures for Performing a Failure Mode, Effects and Criticality Analysis, 1980.
The rating guide and severity flag are practitioner guidance, not a reproduction of any published rating table. Use your organization’s adopted scales for formal FMEAs.