What Reliability Means
The classical definition of reliability, its four parts, and why each part must be stated before reliability can be measured.
12 min
Activities: 0 of 2
A gas export compressor trips three times in one month. Production loses gas, flaring increases, and the operations manager asks a simple question: "How reliable is this machine?" The question sounds simple. It is not. Before anyone can answer it, they must agree what "reliable" means.
Reliability engineering gives a precise answer. The classical definition is: reliability is the probability that an item will perform a required function, under stated conditions, for a stated period of time.
The International Electrotechnical Commission (IEC) vocabulary for dependability, IEC 60050-192, uses a shorter wording: the ability to perform as required, without failure, for a given time interval, under given conditions. Both say the same thing. Reliability is not a feeling about a machine. It is a measurable property, and it has four parts.
The four parts of the definition
Why each part matters in practice
Most arguments about reliability come from a missing part of the definition. The maintenance team says a seawater lift pump is reliable because it "never stops". The process engineer says it is unreliable because its discharge pressure has been low for weeks. They are using different required functions.
In the same way, a vendor may quote a reliability figure for a motor in a clean, cool factory. The motor is then installed on an offshore platform in salt air at 45 °C. The stated conditions have changed, so the vendor figure no longer applies.
When you see or produce a reliability figure, check all four parts. If one is missing, the number cannot be compared with anything else.
A vendor states that a motor is "98% reliable". What is the most important missing information?