Definition(s)
Failure Due to Demand
failure occurring on demand
γ, ψ
failure of one item due to a change of its state triggered by an external event (the so-called “demand”)
EXAMPLE 1 Obtaining 2 when launching a dice is an event occurring on demand. The probability of this event
is 1/6. It does not depend on the elapsing time but only of the demand itself (i.e. the fact that the dice is launched).
EXAMPLE 2 The failure of an electromechanical relay (e.g. rupture of the spring) when it changes state depends
on the number of operations (cycles) rather on the operating time (see IEC 61810–2[49]) and this is the same for
the failure of an electronic device due to over voltage when it is switched or the blocking of a diesel engine when
it is started, etc.: these are typical examples of failures due to demands (or cycles).
Note 1 to entry: In this Technical Report two kinds of demand are considered: the periodic tests and the demand
for an actual safety action. The probability of a failure due to periodic test is a constant number noted γ and the
probability of a failure due to one actual demand of the safety action is a constant number noted ψ. Over a given
time interval, those probabilities of failure do not depend on the duration but on the number of demands or tests
occurring within this interval. The use of γ and ψ is explained in 7.3.
Note 2 to entry: This should not be confused with the “failure on demand” appearing in the term “probability of
failure on demand” (see 3.1.14, Note 2 to entry) used in functional safety standards[2] for low demand mode safety
systems. In those standards this means “failures likely to be observed when a demand occurs”.
Source: ISO/TR 12489:2013(E) Reliability modelling and calculation of safety systems. Global Standards