关键词: hazard identification (HAZID), steel jacket structure, relative importance factor, failure probability, genetic algorithm (GA)

Abstract: This paper proposes a risk-identification-based hybrid method for estimating the system reliability of steel jacket structures under fire. The proposed method starts with risk identification; according to the results of hazard identification and Dow’s fire and explosion index (F&EI) methodology, the most dangerous hazard sources are determined. In term of each equipment layout in steel jacket structures, fire load is imposed and elasto-plastic analysis is performed. According to the deformed state of steel jacket structures, the weakest failure mode of steel jacket structures is identified. In order to know the effect on ultimate bearing capacity of the offshore structural system, a series of elasto-plastic analyses are performed in which single failure element contained in the weakest failure mode is removed from the whole offshore platform structural system. Finally, the failure function of the steel jacket structure is generated and the failure probability of the steel jacket structure system is estimated under fire by genetic algorithm via MATLAB program.

Key words: hazard identification (HAZID), steel jacket structure, relative importance factor, failure probability, genetic algorithm (GA)