How do you calculate an RPN?

Multiply the severity, occurrence, and detection scores together, each rated 1 to 10. A severity of 7, occurrence of 4, and detection of 5 give an RPN that ranks this failure mode against others on the same FMEA.

What is a good or acceptable RPN?

Many teams set an action threshold at RPN above 100 or 150, with anything higher requiring documented corrective action. But a high severity of 9 or 10 alone often triggers action regardless of the multiplied total.

What do severity, occurrence, and detection mean?

Severity rates how harmful the failure effect is, occurrence rates how frequently the cause happens, and detection rates how likely your controls catch it before the customer, where a 10 means you almost never catch it.

Why does a high detection score increase the RPN?

Detection is scored inversely: a 10 means detection is poor and the defect is likely to escape, while a 1 means you reliably catch it. A poor detection score multiplies the risk up because escaped defects reach the customer.

RPN vs criticality: what is the difference?

RPN multiplies all three factors, while criticality analysis focuses on severity and occurrence to gauge consequence and frequency. Criticality is preferred when detection is hard to estimate or when safety severity should dominate the ranking.

How should I act on tolerance stack RPNs?

Sort failure modes by RPN, attack the highest first, and prefer fixes that reduce occurrence (tighter process control) or severity (design margin) over fixes that only improve detection, since catching defects later is the weakest mitigation.

Measurement, Test & Control Equipment calculator

Component Tolerance Stack Risk Calculator

The Risk Priority Number, or RPN, is the core scoring output of a Failure Mode and Effects Analysis, and here it is applied to component tolerance stack-up risk. Quality engineers, metrology teams, and design-for-manufacturing reviewers use it to rank which dimensional failure modes deserve corrective action first. By multiplying how bad a failure is (severity), how often it occurs (occurrence), and how likely you are to catch it before it ships (detection), you collapse three judgments into a single comparable number from 1 to 1000. For tolerance stacks specifically, it flags where accumulated part-to-part variation is most likely to push an assembly out of spec without being caught at inspection.

What this calculator does

Score the risk of tolerance stack-up failures in assemblies that affect measurement accuracy or test equipment performance. Uses severity, occurrence, and detection scoring similar to FMEA methodology.
Use when evaluating tolerance stack risk in precision assemblies, ranking measurement system design risks, or prioritizing design changes during FMEA reviews for test fixtures and instrument assemblies.
It computes the FMEA Risk Priority Number by combining severity, occurrence, and detection scores into one ranked risk figure.

Formula used

Risk priority number (RPN) = severity score x occurrence score x detection score
RPN range: 1 (lowest risk) to 1000 (highest risk). Typical action threshold: RPN > 100 or 150.

Inputs explained

Severity score: Rate the impact if tolerance stack causes a failure: 1 = no noticeable effect, 5 = reduced measurement accuracy or increased uncertainty, 10 = complete loss of measurement capability or safety hazard.
Occurrence score: Rate how often worst-case stack occurs: 1 = extremely unlikely (all suppliers Cpk > 2.0), 5 = occasional (Cpk around 1.0), 10 = frequent or near-certain based on process capability data.
Detection score: Rate your ability to detect the problem before it reaches the customer: 1 = 100% functional test catches it, 5 = sample inspection or dimensional audit, 10 = no current inspection or test covers this failure.

How to use the result

Use it during design reviews, PPAP, or process audits to prioritize which tolerance-related failure modes get corrective action.
RPN is ordinal, not a probability, and equal RPNs can hide very different risk profiles, so always review the underlying severity score independently.

Current U.S. benchmarks

Steel mill PPI stands at 348.53 (BLS, May 2026), up 6.7% from a year earlier. New factory orders are up 2.3% year over year (Census).

Common questions

How do you calculate an RPN? Multiply the severity, occurrence, and detection scores together, each rated 1 to 10. A severity of 7, occurrence of 4, and detection of 5 give an RPN that ranks this failure mode against others on the same FMEA.
What is a good or acceptable RPN? Many teams set an action threshold at RPN above 100 or 150, with anything higher requiring documented corrective action. But a high severity of 9 or 10 alone often triggers action regardless of the multiplied total.
What do severity, occurrence, and detection mean? Severity rates how harmful the failure effect is, occurrence rates how frequently the cause happens, and detection rates how likely your controls catch it before the customer, where a 10 means you almost never catch it.
Why does a high detection score increase the RPN? Detection is scored inversely: a 10 means detection is poor and the defect is likely to escape, while a 1 means you reliably catch it. A poor detection score multiplies the risk up because escaped defects reach the customer.
RPN vs criticality: what is the difference? RPN multiplies all three factors, while criticality analysis focuses on severity and occurrence to gauge consequence and frequency. Criticality is preferred when detection is hard to estimate or when safety severity should dominate the ranking.

Last reviewed 2026-05-12.