Definition
Failure rate expresses how frequently failures occur, per unit of time (or cycles), for a defined population of assets and a defined observation period, for example, failures per 1,000 operating hours. It is closely tied to MTBF (often its mathematical reciprocal for constant-rate periods) but is expressed as a rate rather than an average interval.
Key facts
- Expressed as failures per unit of time or cycles for a defined population, not a single asset's individual history alone.
- Mathematically related to MTBF (approximately its reciprocal during the constant-failure-rate period of an asset's life), but the two are read differently.
- Varies over an asset's life according to the bathtub curve: early life, useful life, and wear-out phases typically show different rates.
- Requires a consistent population definition and observation window to be comparable over time.
How is Failure rate calculated?
Failure rate is calculated by counting the number of failures observed across a defined population of assets during a defined observation period, and expressing that count relative to the total operating time or cycles accumulated, for example, 3 failures per 10,000 operating hours across a fleet of pumps.
What influences Failure rate?
MTBF expresses the same underlying reliability information as an average interval between failures. Failure rate expresses it as a frequency. The bathtub curve describes how failure rate itself typically changes over an asset's life (higher during early-life and wear-out phases, roughly constant during useful life) which is why a single average failure rate can be misleading if the asset's life phase is not considered.
FAQ
Frequently asked questions
How is the frequency of failures related to a reference quantity?
Failure rate: the number of failures observed for a defined population of assets, expressed per unit of time or operating cycles over a defined observation period.
Last reviewed: