Every day, management and maintenance teams are either being praised or blamed because of decisions they made with regard to maintenance strategies used in a plant.
Reactive maintenance can provide low operational costs. Yet in the worst case, this strategy can also wipe out any cost efficiency if a plant’s uptime is jeopardized. Predictive maintenance requires a completely different operational culture. It can even be easily neglected, although this strategy may provide the best plant availability.
By understanding the benefits and weaknesses of each of the three types of maintenance strategies, it is possible to make the right decision of which to use in every case. In fact, especially for new projects, it is possible to use a combination of all three strategies to cover even on/off valves without any considerable extra costs.
Reactive maintenance is essentially letting something run until it breaks. This means that no effort or action is taken to maintain the equipment. This was the most predominant maintenance mode in the US as late as 2000.
Preventive maintenance is carried out according to the calendar year. This is a good approach for standard replaceable components that wear out with use.
Predictive maintenance takes advantage of diagnostics information to help make a better maintenance plan. Diagnostic-based maintenance generally shortens repair time and extends the useful life of the asset.
Introducing Reliability Centered Maintenance – RCM
RCM recognizes that not all equipment within a plant is of equal importance to the process or its safety. Therefore, the assets are classified based on their criticality and the maintenance strategy is selected accordingly. RCM suggests that less than 10% should be allocated to reactive maintenance, 25–35% to preventive maintenance and 45–55% to predictive maintenance. When comparing the different strategies, management may not readily understand the savings potential due to startup costs.
Still, it’s important to remember that predictive and RCM can be enabled easily when the assets are intelligent. The steps are simple. Select assets capable of analyzing and storing performance information during the normal plant run time. Analyze performance in the real operating conditions. No extra work is needed to carry out tests – just simply analyze the results.