Modeling Knowledge Retention in PRA Models for Mission Critical Infrastructure

SUMMARY & CONCLUSIONSKnowledge retention is an issue that impacts the risk of failure of large, complex systems over their typically very long lifecycles. Attrition, retirements, layoffs, and generational turnover all lead to diminishing knowledge in the workforce [1]. Risk assessment methods, such as probabilistic risk assessment (PRA), typically do not account for knowledge retention specifically [2].This paper proposes a method of analyzing the risks that a complex system faces as the operators and organization collectively lose knowledge over time. The method is presented and demonstrated on a fictionalized civilian nuclear power plant (NPP) that explicitly adds knowledge retention as a basic event (BE). The Standardized Plant Analysis Risk Human Reliability Analysis (HRA) (SPAR-H) [3] method is extended to account for this by including knowledge retention BEs under OR logic gates in the PRA fault trees (FTs).A case study of a fictionalized civilian NPP is presented to demonstrate the proposed method. This paper demonstrates that knowledge retention should be accounted for in PRA models to ensure that a complex system's failure rate is more adequately represented over the lifecycle of the system. The proposed method may help organizations forecast the need for increased training, accelerated hiring, or improved retention policies to ensure adequate staffing with sufficient knowledge to safely operate complex systems across their entire lifespans.