Mitigating production losses and reliability analysis via RCM-FMEA driven instrumentation maintenance program

Abstract Production loss represents a significant hazard in the process industry, and when coupled with catastrophic events, the financial consequences can be severe. High equipment failure rates and the associated risks of lost production time underscore the need for an optimal maintenance program that ensures safe, reliable, and extended operational hours for equipment or plants while minimizing the risk of production loss. Industrial plant maintenance has evolved gradually with time. Starting with well-known reactive maintenance (RM) techniques, such as run-to-fail (RTF) and emergency maintenance (EM), which implies a corrective action after the equipment failure. Further evolution led towards proactive approaches of preventive maintenance (PM)—a time-based maintenance approach, to condition-based maintenance (CBM)—monitoring of current equipment parameters in comparison with defined thresholds, to predictive maintenance (PdM)—a failure prediction approach based on statistical data analysis or with the support of AI-driven platforms. The latter three historical approaches, PM, CBM, and PdM, have proven records of reducing plant downtime; however, each has its own pros and cons regarding manpower resource allocation and cost. Developing an effective, optimal maintenance program has always been a challenge. A historical technique named Failure Mode and Effect Analysis (FMEA), in combination with Reliability Centered Maintenance (RCM) , has shown promising results in decision-making, evaluation and selection of an appropriate maintenance task once applied on earlier mentioned strategies and adding a significant value to the Maintenance program in terms of effectiveness while addressing each failure mode with a systematic approach and as a resultant enhancing the plant overall reliability at optimal performance levels with an optimal effort and avoiding critical equipment failures and mitigating production losses.