Objectives. A necessity for this study was felt in the catalyst replacement process as a maintenance operation, because some fatal incidents have occurred due to human error in process industries during catalyst replacement operation. Identification and evaluation of human error is essential in predicting and reducing accidents in maintenance operation. Methods. The human error assessment and reduction technique (HEART) as a human reliability analysis was applied in this study. Because the HEART method uses expert judgment, it is possible to make a bias in the assessment. Fuzzification and aggregation opinions of a heterogeneous expert group in a trapezoidal fuzzy set were used to mitigate this deficiency. Results. Evaluation results showed that the subtask 'Entering the reactor' with human error probability (HEP) of 9.2E-1 and the subtask 'Reduce temperature while reducing feed' with HEP 1.3E-3 had the highest and lowest chance of human error, respectively. Conclusions. In order to reduce the likelihood of human error, it is necessary to take appropriate actions based on the task-specific error producing condition (EPC).
Keywords: catalyst replacement; fuzzy logic; human error; human error assessment and reduction technique; maintenance.