The Complimentary Roles Of Reliability-Centered Maintenance and Condition Monitoring By Richard Overman, M.S., CMRP Chief RCM Engineer and Roger Collard, CMRP Reliability Engineer Advanced Information Engineering Services (formerly Veridian)

Originally presented at IMC-2003, the 18th International Maintenance Conference

Click here for a print friendly 66k pdf version

ABSTRACT

Condition Monitoring (CM) technologies, such as vibration analysis, infra-red thermal imaging, and ultra-sonic flaw detection along with many others have been widely used for detecting eminent equipment failures in various industries.  Some applications have been more successful than others.  While CM technologies are founded in sound scientific practice and usually display proven track records, their application must be administered through best business practice and value added efforts.  The Reliability Centered Maintenance (RCM) methodology lends itself fully to the application of the correct maintenance procedure being conducted at the correct time and by the correct person(s).  The reader will discover how the RCM analysis not only advises the administration of the correct technology, but also the acceptable interval for the inspection.  Also discussed, are the economic considerations of the CM technologies as part of the RCM analysis, allowing the user to investigate the level of expertise required to produce acceptable results for a given technology. 

It seems that each year, large capital investments are made into the latest technologies, with minimal thought given to an effective applications program.  Often times, this, results in a substandard return on investment.  This paper depicts the increased effectiveness of CM technologies when integrated with an appropriately developed RCM analysis.  Such integration avoids duplication of effort and misapplication of CM technologies, ensuring inspection efforts are directed at pre- determined failure mode and consequence mitigation.

INTRODUCTION

When I think of the integration of RCM and CM technologies, two conversations come to mind.  While both of these conversations involved vibration analysis, they could have as easily involved any of the other CM or predictive maintenance technologies.

The first conversation involved a plant that used large fans during their production process.  While talking with one of the maintenance engineers, he mentioned that they use vibration analysis to tell them when the fans need cleaning.  Later in the conversation, he stated that they shut the fans down every 6 months to clean them.  The obvious question is, why do both?  If the vibration analysis adequately predicts when the fans need to be cleaned, why have a scheduled shutdown just to clean them?  On the other hand, if the plant is already being shut down, and cleaning the fans at that time will let them operate for another 6 months, why do vibration analysis?  The RCM analysis process is designed to provide a well documented, structured way to evaluate these and other function preservation strategies.

The second conversation involved a plant that uses a lot of pumps.  The plant maintenance personnel stated that they perform vibration analysis on the pumps once a month.  When asked why, the response was that this is what the vibration equipment vendor recommended.  Upon further questioning, the technicians revealed that the pumps would run about 6 months after a potential problem is detected by vibration analysis.  A reasonable question in light of the 6 month warning is, why perform vibration analysis every month, why not every 2 or 3 months?  An additional piece of information is that the plant requires 100 percent pump availability in the summer months, but can accept pump failures during the winter months.  This raises the question, why do vibration analysis in the winter months at all?

The key point is that CM technologies are often treated as an end product rather than one of many possible function preservation tools.  The RCM analysis process provides a well-documented, structured method for evaluating the efficient, effective use of CM technologies. 

CONDITION-MONITORING TECHNOLOGIES 

A plethora of condition-monitoring technologies have sprung up over the last 40 years in response to a specific need.  This need was revealed in a study performed for the commercial airline industry in the early 1960’s during the development of the preventive maintenance program for the “new” Boeing 747.  “The FAA initially envisioned this program to be 3 [times] more extensive than the 707 program under the rationale that the 747 would carry 3 [times] more passengers.”[1]  The airlines knew that such a program would not be economically viable and launched a major study to validate the failure characteristics of aircraft components.  Figure 1 shows the results of that study.  

In figure 1 we see that only 11 percent of the components demonstrated a failure characteristic that supported a scheduled overhaul or replacement (scheduled removal).  Eighty nine percent exhibited random failure characteristics for which a scheduled removal was not effective.  Since a scheduled removal was the primary scheduled maintenance program at that time, new ways were needed to deal with the 89 percent not applicable to scheduled removal.  Enter CM technologies.  CM technologies were developed to predict the onset of failure for components that exhibit a random failure characteristic.

Figure 1- Aircraft Component Failure Characteristics[2]

Over the years, various names have been given to the family of CM technologies, such as on-condition maintenance, condition based maintenance, preventive maintenance, and predictive maintenance.  New CM technologies are continually being developed with increasing focus on computerized systems to perform continuous CM of equipment.  For the purposes of this paper, a CM technology is one that checks the condition of the component or process on a regularly scheduled basis to look for the onset of failure.  Regularly scheduled can be measured in terms of months, operating hours, or microseconds, as would be the case for “continuous monitoring” tasks.

Over the years much attention has been devoted to the development of these technologies.  A reasonable question is whether more attention needs to be given to the efficient and effective use of these technologies.  To do this we need to ask, how does one decide which of the many CM technologies to use?  How often should a CM technology be applied?  Is continuous CM worth the investment?  How good are CM technologies at detecting the onset of failure?  These and other questions need to be addressed for specific equipment within the specific application of the equipment.  The RCM process provides a well-documented, structured process for evaluating these questions.

RELIABILITY-CENTERED MAINTENANCE

The study that spawned the development of CM technologies also spawned the development of the RCM process.  To identify the appropriate maintenance requirements for the Boeing 747, representatives of various airlines developed a process that became known as the maintenance steering group (MSG) logic.  In 1978, the Department of Defense asked Stanley Nolan and Howard Heap, both from United Airlines, to expound upon MSG philosophies for application to military aviation.  Their report coined the name “Reliability-Centered Maintenance”.  After MSG-1, RCM development followed three distinct and separate tracks as shown in figure 2.  The three tracks are the commercial aviation track, the military aviation track (led by the Navy) and the commercial industry track.  The commercial industry track became the most diverse track with many different groups and people entering the market.  RCM became divided into 2 main groups; the “classical” RCM processes and hybrid RCM processes.  Hybrid RCM includes various attempts at taking short cuts with the RCM process, usually by leaving out some steps.  The Society of Automotive Engineers (which involves every mode of transportation including rail, aviation, automobiles, and space) saw a need to write a standard[3] that defines what a process should include in order for it to be a “true” RCM process – that is, a process that conforms to the original RCM concept and one that includes all of the steps necessary to keep from being dangerous.  This standard was published in 1999, and this author had the honor of serving on the committee to write the standard.

The SAE standard defines RCM as, “A specific process used to identify the policies which must be implemented to manage the failure modes which could cause the functional failure of any physical asset in a given operating context.”  It can also be looked at as a process for evaluating function preservation strategies.  The goal of the RCM process is to ensure that the right people perform the right maintenance, at the right time, in the right way, with the right training and tools.

Figure 2- History of RCM

CM is one of the many function preservation strategies evaluated during an RCM analysis.  The other function preservation strategies are other scheduled maintenance, design changes, training improvements, operational changes, on-time changes, and run-to-failure.  The remainder of this paper addresses the RCM evaluation of the application of CM technologies.

PUTTING CM AND RCM TOGETHER

 At the end of the earlier CM discussion, four key questions were asked.  They were:

1.       How does one decide which of the many CM technologies to use? 

2.       How often should a CM technology be applied? 

3.       Is continuous CM worth the investment? 

4.       How good are the CM technologies at detecting the onset of failure?

RCM provides the analytical philosophies to effectively answer these questions. 

The RCM process addresses the first two questions in the determination of the degradation interval (see figure 3).  Figure 3 is rather complicated, and explaining it in detail is beyond the scope of this paper.  There are a few main points germane to this paper.  The RCM standard gives the following criteria for the technical feasibility of a CM task[4].

1.       “There shall exist a clearly defined potential failure” (point B on figure 3).

2.       “There shall exist an identifiable P-F interval” (P-F stands for “potential to functional failure” interval and is the same as the degradation interval on figure 3, interval from point B to point C).

3.       “The task interval shall be less than the shortest likely P-F interval” (inspection interval < degradation interval).

4.       “It shall be physically possible to do the task at intervals less than the P-F interval.”

5.       “The shortest time between the discovery of a potential failure and the occurrence of the functional failure (the degradtion interval minus the task interval) shall be long enough for predetermined action to be take to avoid, eliminate, or minimize the consequences of the failure mode.”

During the RCM process, the technically feasible criterion is applied to all CM technologies that might be used to detect the potential failure.  An inspection interval is identified for each one.  The inspection interval may be different because different CM technologies can detect the onset of failure at different places along the degradation curve.  For safety and environmental consequences, the task is technically feasible if the task at the identified interval reduces the probability of failure to a defined tolerable level.  For operational and non-operational consequences, the task at the identified interval is technically feasible if it is cost effective.  A cost analysis is performed on all of the technically feasible technologies to see which is the most cost effective. 

One of the technically feasible options may be to install a system that continuously monitors the asset to identify the potential failure condition as soon as possible.  A cost analysis of this option along with the other options is part of the RCM process.  The continuous monitoring option is the one worth doing if it is the most cost effective.

 Figure 3- Degradation Interval

Finally, a word about the fourth question listed previously.  Part of identifying the inspection interval involves an estimate of the task effectiveness.  This is usually addressed as the probability that a potential failure will be found, assuming that it exists.  This is a part of the overall CM technology development that needs more attention.  Academic studies into how effective CM technologies are at various levels of expertise are in order.  A great deal of work has been done on developing the technologies; however, more work needs to be done on how effective they really are.

CONCLUSION

CM technologies are excellent tools for identifying potential failures.  As tools they need to be applied in the right circumstances and only when necessary.  The RCM process is designed to identify those circumstances and to determine when they are necessary.