Performance standard

A performance standard turns a critical control from an assertion into a testable contract. At minimum it states: the control's objective (what it must achieve, stated as an outcome), the criteria that show it is achieving it, the verification that checks those criteria — both system checks and field checks — and the trigger: what happens when verification finds the control below standard.

The concept comes from safety-case regimes (UK offshore practice defines performance standards for safety-critical elements in terms of functionality, availability, reliability, and survivability) and from ICMM critical control management, which asks the same questions of any critical control: what does good look like, who checks, how often, and what do they do when it isn't good?

Two habits keep performance standards useful rather than ornamental. Write criteria that a verifier can fail — "RBI intervals current for every circuit" can be failed; "inspection programme in place" cannot. And record how each criterion erodes, because that links the standard to the degradation factors the bowtie already names.

In Bowtie Risk Engine the performance standard is authored on the control itself and exported in the audit PDF's annex — every control in the process safety example carries one, if you want to see the shape in practice.