Improved Z-number and Bayesian network analysis to predict accommodation electrical fire risk on a tanker ship

Accommodation fire risk is of paramount importance and highly dangerous for cargo ships, particularly tankers that carry hazardous liquid cargo. The consequences of an accommodation fire may be fatal since the tankers carry flammable liquid cargoes. This paper aims to assess the potential risk of accommodation electrical fire on a tanker ship. To achieve this objective, the paper adopts a Bayesian network (BN) together with improved Z-numbers theory. In the model, BN enables a robust tool based on a cause-and-effect network between the variables, and the improved Z-numbers are able to cover uncertainty and improve the reliability of the analysis. In view of the findings, the occurrence probability of accommodation electrical fire risk on a tanker ship is found to be 9.60E-02. The research outcomes will significantly contribute to understanding the causal network of the electrical fire, as well as enhancing safety.