NSGI

Reliability and Asset Management

Archive for the ‘Maintenance Basis & RCM Analysis’ Category

Optimizing Preventive Maintenance and the Maintenance Basis

Thursday, December 2nd, 2010 | Maintenance Basis & RCM Analysis | 43 Comments

Have you “Optimized” your PM program only to end up with more preventive maintenance tasks than you started with? – If you have, you are not alone; the majority of the organizations that we have talked with ended up with more to do not less and they seriously questioned the value of the new task in their PM program.

When optimizing a preventive maintenance program there are two key elements to keep in mind.   The first is to effectively classify equipment and the second is to place an emphasis on predictive maintenance and equipment condition monitoring tasks.  Effective classification of equipment allows the proper focus of resources for both the optimization and the conduct of the Preventive Maintenance program.  Effective equipment classification should be based upon a clear understanding of the probability of equipment failure and the consequences of that failure, so as to minimize the risk to safety, production, and operating budgets. Historically many plants have used a 2 tier system for classifying equipment.  The equipment was classified as either critical or non-critical with all safety related equipment having a default classification as critical.  In theory the most plants in the industry now uses a minimum 3 or 4 tiered system.  However, many plants still seem to default to the 2 tier system, “critical and everything else”, due to an excessive number of PMs.  When optimizing a PM program a six tiered priority system seems to provide enough granularity to effectively allocate resources, the exact number should be determined by your plant.  A caution should be noted here, use of a large number of tires in the classification systems risks the chance that the system becomes too complicated to be useful.  In that case the organization ends up defaulting back to a 2 tiered system.  The following is an example of an equipment classification system for consideration:

  • Safety Critical – functional failure could result in significant injury to plant personnel or the failure of a nuclear safety function
  • Production Critical – functional failure could result in all or major loss of production capability
  • Safety Important – functional failure require significant effort to put in place barriers to protect personnel
  • Production Important – functional failure could result in a significant loss of production capability
  • Minor – functional failure has little or no impact to production or safety however it is financial prudent to perform work prior to failure
  • Run-to-Failure

Adhering to the classification system that is developed is important and will take organizational discipline.

Most of us are familiar with the “bathtub curve” that was developed out of equipment reliability studies performed in the 1950’s.  This curve has us make the assumption that as a whole our plant equipment has some period of infant mortality, followed by a period where the probability of equipment failure is fairly flat and then close to the end of the equipments usefully life the probability of equipment failure increases.  If we follow this assumption, our natural tendency then is to develop maintenance tasks timed to be performed just before a piece of equipment would fail to perform its intended function; thereby allowing for the longest period of useful life and the lowest probability of production interruptions due to equipment functional failures. If equipment failures did occur, the “bathtub curve” might also draw us to the conclusion that there were either an insufficient number of PMs or that their frequency of the existing PMs was insufficient to prevent the equipment failure.   This is how most early PM program analysis was performed and PM programs grew to be too large to effectively manage.

During the development and refinement of Reliability Centered Maintenance (RCM) analysis methodology Stanley Nowlan and Howard Heaps broke down the “bathtub curve” into six separate curves.  Their studies showed that only 11% of equipment failures followed an age related degradation path while 89% of equipment failures were random in nature.

Nowlan & Heap

While some reliability experts may argue over the exact percentages of age related failures versus random failures the general consensus is that random failures far outnumber age related failures in our plants.  So why then, are the majority of plants maintenance strategies predominately comprised of time based maintenance (preventive maintenance) tasks?  If the majority of equipment functional failures in our plants are random in nature, then shouldn’t the majority of maintenance strategy tasks be designed to detect the onset of those random failures?

Predictive Maintenance and Condition Monitoring (or CBM) tasks are designed to identify degrading conditions in equipment prior to functional failure.  The appropriate PdM/CBM dependent on the type of equipment and failure modes associated with the equipment and its operating environment.  The frequency at which PdM/CBM tasks should be performed should be based upon a number of factors including; the risks (probability and consequence) associated with functional failure, the availability of operating conditions needed to collect data, the ability of the monitoring technique(s) being applied to detect the onset of degradation, the skill of the technicians collecting and analyzing the data, and the rate of degradation of the of the failure modes being evaluated.

So in summary, an effective PM optimization effort, along with all the other considerations management may place on the effort, must stress the accurate classification of the equipment and provide a focused effort on replacing time based tasks with PdM/CBM tasks.

Tags: , , , , , ,

Copyright © 2024 NSGIWebsite design by Slamdot.