In brief
What to carry into the engineering work
- Write requirements that are specific, testable and traceable to the hazardous scenario.
- Keep functional behaviour, integrity targets and lifecycle controls distinct but connected.
- Control document identity, assumptions, review status and changes as part of the engineering record.
Document identity and basis
Start with the project, unit, SIF tag, document number, revision, status, preparer, checker and approver. Link the hazardous scenario, risk-study reference, required SIL and assumptions that created the requirement.
Functional requirements
- Initiating condition and trip setpoint.
- Safe state and required final action.
- Operating modes in which the function is required.
- Process safety time and maximum SIF response time.
- Reset, restart, latching and manual-action philosophy.
- Interfaces with BPCS, utilities and other protection layers.
Integrity and architecture
State the demand mode, required SIL, quantitative target, applicable architectural route, systematic-capability requirement and any independence constraints. Reference the verification method rather than embedding a changing calculation result as a requirement.
Testing, bypass and maintenance
Define proof-test interval and required coverage, diagnostic response, maximum restoration time, bypass authorization, compensating measures, alarm handling and records. A requirement should be testable and have a clear acceptance criterion.
Review and issue
Resolve assumptions or mark them as controlled actions before issue. Check consistency with the cause-and-effect, logic diagrams, instrument data sheets, proof-test procedure and verification record. Validation should trace back to the issued requirements.
Primary references
- IEC 61511-1 — process-industry SIS framework and requirements
- UK HSE — Functional safety
- ISA — ISA-84 series of standards
Use the applicable edition and national adoption. Project specifications and controlled procedures may add requirements.