Preventive Maintenance Optimisation: an Application to Critical Process Instrumentation in a Petrochemical Plant

It is important in any process industry to keep under control product quality, human and environmental safety, and operating costs. The effectiveness of controls depends on the reliability/availability of accurate measuring instrumentation; this can only be achieved through correct maintenance strategies. Results of risk analysis on petrochemical plants show that the calibration failure mode has the highest occurrence. Unfortunately the arrangement of calibration service is still based on practical considerations of maintenance operators (often arbitrary) and on the companies need for limiting operational costs. In order to optimise the calibration intervals, it is possible to analyse calibration history data collected during past maintenance operations on the instruments. A statistical method was developed to dynamically evaluate the optimal time interval for the calibration service, and to assess the plant safety level and the capability of the maintenance service through synthetic indices.