Using risk data as a source for human reliability assessment during shipping LNG offloading work

Shipping Liquefied Natural Gas (LNG) has become a popular method for transporting LNG. However, offloading work poses significant risks, many of which are attributed to human errors. Considering that most accidents are associated with human errors, the Human Reliability Analysis (HRA) method is a critical option to prevent accidents and to predict Human Error Probability (HEP). One such method is fuzzy CREAM, a well-known HRA methods. However, this method has some limitations. The method uses Common Performance Conditions (CPCs) to estimate HEP, but the source of CPC data is insufficient. Without enough and reliable CPC data, the process and the result of fuzzy CREAM are questionable and criticized. Therefore, this study proposes a modified approach to address this issue. The proposed method uses the definition of “Risk” as the support to collect each CPC's data from aspects of likelihood and impact. Then, using the collected risk data as the source to determine each CPC's fuzzy degree, to determine each CPC's weight by combining with Grey Relationship Analysis (GRA), and to identify each activated fuzzy If-Then rules and the rule weight. Afterwards, the proposed method integrates the fuzzy degree of each CPC, the weight of each CPC, and the weight of each activated If-Then rule together to estimate HEP. Finally, the proposed method is validated through a real engineering case of shipping LNG offloading work.