Functional safety technical resource
What Is a Risk Graph?
A risk graph is a structured qualitative or semi-quantitative method for determining a required Safety Integrity Level (SIL) or risk-reduction category by following a predefined path through calibrated risk parameters. Common parameters address consequence severity, frequency or duration of exposure, possibility of avoiding the hazardous event, and probability of the unwanted occurrence. A risk graph is a decision aid, not a universal calculator: its result depends on the organization's calibration, parameter definitions, scenario framing, and treatment of existing safeguards.
Technical resource from Mangan Safety Instrumented Systems (Mangan SIS)
How the decision path works
The analyst selects one category for each parameter using the definitions in the approved procedure, then follows the corresponding branches to an output. The scenario should identify the hazardous event, consequence endpoint, exposed population, operating mode, initiating conditions, and safeguards considered before the graph is used. Each selection needs a recorded rationale; an unexplained letter code or final SIL is not an auditable risk assessment.
Calibration gives the graph meaning
The category boundaries and branch endpoints are calibrated to the organization's tolerable-risk criteria and assumptions about existing protection. A graph copied from another company, textbook, or project can produce an unsupported result because consequence definitions, exposure ranges, avoidance credit, occurrence categories, and risk targets differ. Calibration should also address how multiple fatalities, environmental harm, and asset criteria are handled without combining unlike consequences casually.
Judgment and boundary sensitivity
Risk-graph categories compress uncertainty into discrete choices, so scenarios near a boundary can change output when one category changes. Avoidance and exposure judgments can be especially sensitive to occupancy, warning time, escape routes, event speed, and operating mode. The team should document uncertainty, test plausible adjacent selections, and escalate a decision when the output is highly sensitive or the scenario does not fit the graph's calibration.
How a risk graph differs from a risk matrix
A risk matrix usually combines consequence and likelihood to rank risk or prioritize action. A risk graph follows a prescribed multi-parameter path intended to map a hazardous scenario to a risk-reduction or SIL category. Neither method proves the achieved performance of a SIF. After selecting a required SIL, the team must specify the function and independently verify its architecture, random-hardware performance, systematic integrity, response, testing, and lifecycle controls.
Mangan SIS project context
Apply Risk Graph to documented SIS decisions
Use a risk graph only under a controlled procedure whose parameter definitions, calibration, facilitator competence, safeguard treatment, and documentation requirements are approved. Apply it scenario by scenario, record dissent and uncertainty, and use a more quantitative method such as LOPA when the graph is not well matched or the decision is sensitive.
Sources and scope
Use the source type that fits the decision
Standards
IEC 61511-1 requires the required SIL of each SIF to be determined from hazard and risk assessment, but it does not require one universal risk-graph design. IEC 61511-3 provides informative examples and guidance for SIL-determination methods rather than prescribing the SIL for a particular application.
Guidance
IEC 61508-5 illustrates methods for determining safety integrity levels and explains underlying risk concepts. Risk-graph examples are simplified to show principles; users must establish application-sector and organization-specific criteria rather than treating an example graph as a ready-made standard threshold.
Regulation
OSHA 29 CFR 1910.119 requires process hazard analysis for covered processes but does not require a risk graph, assign SIL, or define risk-graph category boundaries. A graph can support an engineering decision only if the overall PHA addresses the applicable regulatory elements.
Mangan recommendation
Mangan SIS recommends governing one calibrated risk-graph procedure with version control, category examples, safeguard-credit rules, competence requirements, and escalation criteria. Preserve the selected categories and rationale, and challenge boundary-sensitive results with an alternate method before final SIL allocation.
Primary references
- IEC 61511-3:2016 – guidance for required SIL determination (International standard guidance)
- IEC 61508-5:2010 – examples of SIL determination methods (International standard guidance)
- ISA-84 Series – process-sector functional-safety standards (Official standards catalog)
- OSHA 29 CFR 1910.119 – Process Safety Management (U.S. regulation)
Common questions
Clarifying Risk Graph
What inputs does a risk graph use?
Common graphs use categories for consequence, exposure frequency or duration, possibility of avoiding the hazardous event, and probability of the unwanted occurrence. The exact meanings and boundaries must come from the organization's calibrated procedure.
Is a risk graph the same as a risk matrix?
No. A risk matrix generally combines consequence and likelihood to rank risk. A risk graph follows a multi-parameter decision path intended to map a scenario to a required risk-reduction or SIL category.
Does a risk-graph result prove that a SIF achieves its SIL?
No. The graph can help select the required SIL. The complete SIF must then be specified, designed, verified, validated, operated, tested, maintained, and managed to show that the required performance and lifecycle conditions are met.
Selecting a required SIL?
Use calibrated criteria, documented assumptions, and an auditable link to the hazard scenario.
Downloadable term reference
Download: Risk Graph – Functional Safety Glossary PDF
Save or share this term-specific reference, including the answer-first definition, engineering review considerations, and primary-source references.