An algorithmic tool for supporting risk and root-cause analysis of critical incidents in Baltic Sea Region ports

This paper is both a summarization and extension of [6] and [7], where a stochastic model of interacting operations carried out in a generic Baltic Sea Region port was proposed and analyzed. Each operation involves a number of possible unwanted events (critical incidents) whose instances occur randomly and can cause instances of other events affecting this or other operations. This can lead to a cause-effect chain of events affecting one or multiple operations. The model presented in [6] is somewhat complex, therefore it was downgraded to a simpler, application-oriented version demonstrated in [7], where an algorithm computing the risks of critical incidents is constructed and then applied to a real-life example. The current paper, apart from presenting a method of computing the risks of critical incidents, occurring by themselves or resulting from the cascade effect, also features a method of root-cause analysis of such incidents. First, the formulas for the root-cause probabilities are derived, where such a probability quantifies the likelihood that a critical incident occurring in step h of a cascade was caused by another incident that initiated this cascade. Second, an algorithm computing the root-cause probabilities, based on the derived formulas, is constructed. This algorithm is illustrated by its application to the example given in [7]. The presented results can be used as a tool for fault propagation analysis and fault diagnosis applied not only to a port environment, but to any complex industrial system.

---

Opracowanie rekordu w ramach umowy 509/P-DUN/2018 ze środków MNiSW przeznaczonych na działalność upowszechniającą naukę (2019).