A Markov-based model to mitigate human errors occurrence during maintenance activities in petrochemical systems

Human errors aren’t uncommon in maintenance operations to disrupt the scheduled maintenance tasks with adverse consequences on equipment, personnel, or environment. Yet, human error can’t be simply detected during maintenance operation (only after the commissioning). The consequences for such unintended errors may be drastic in petrochemical industries especially under severe operation conditions. The purpose of this study is to explore root cause analysis for human errors in Preventive Maintenance (PM) activities in petrochemical industry. The paper collects the data of poor maintenance incidents from ERP system in one of the largest oil and gas facilities in Saudi Arabia. The incidents analysis indicates 103 failure caused by human errors during the analysis periods from year 2010 till 2019. The findings show that lacking in quality assurance and quality control program is the main cause to have repetitive human failures in PM tasks. Accordingly, a Markov-based model is developed to simulate the improvement effect of having quality assurance and quality control programs as part of the PM tasks. Interestingly, the availability of the equipment is improved by 20% with the quality assurance program and the probability to have human errors dropped to less than 5%.