Does anybody have experiance of the RCM type process used in the PWR nuclear stations in the US. A bit of background on its effectiveness would be much appreciated.
I have an extensive understanding about AP 913 being one of those responsible for helping to develop it. You can reach me at firstname.lastname@example.org.
Author, "Reliability Centered Maintenance- Implementation Made Simple" published by McGraw-Hill
Warning. What follows is a little bit of a rant. Not directed toward the original poster. Just wanted to share my views on RCM and the nuke industry.
I have worked in the nuke industry my entire working life. I guess that leaves me at a disadvantage to compare us to other industries. But I will tell you my perception. My perception is that compared to most other industries (for example fossil-fueld power generation), nukes have a more sophisticated approach to managing equipment reliability towards higher levels. And I suspect they achieve higher reliability.
And by and large AP-913 describes what nukes do.
Can you put together the previous fact to conclude that AP-913 is a silver bullet to achieve reliability? IMO, no way.
For one thing, nukes have a different balance of cost/reliability than others, and typically have a lot bigger engineering and maintenance staff relative to the amount of plant equipment.
For another thing, I honestly don't think there have been many revolutionary ideas in "RCM" (or pick your own favorite in-vogue TLA or buzzword) in a very long time (with the possible exception of new possibilities that increasingly become available with advances in computer technology).
If you have worked in industrial plants for any length of time, the principles of RCM should be no secret.
What is the challenge IMO is not to discover the principles of RCM (those are well known.) What is the challenge is to implement them effectively. That includes having the well-conceived programs matched to the organization, having the discipline to follow the programs in the long term (in spite of numerous competing short term demands), having the self-awareness to recognize when the programs need to evolve, and providing the necessary resources to support those programs.
Nuke plants have one or two advantages in this regard. One advantage is that we have a strong culture to follow procedures. (Our programs are embedded into procedures). This helps with the discipline part. Another part is that we have those resources (mostly people). Yet another advantage is that we are subject to ongoing assessments from internal and external organizations to gage our effectiveness and compare us to others within the industry (most notably INPO assessments). This helps keep us in a self-critical mindset where we can recognize our weaknesses and make appropriate adjustments.
If you took the AP-913 program and slapped it on top of a fossil plant without providing all those other things (starting with the resources and culture/disipline to follow the program) of course it would not achieve the same results as in a nuke plant. It would be viewed as overly complex and cumbersome and would never be accepted. But it might be a good starting point perhaps for training if the trainee is not familiar with RCM. The flowchart describes various things that could/should be considered in the different phases of the "work process" (long-range strategizing, short range work scheduling, work implementation process, and post-implementation and failure/analysis steps) with many many feedback loops.This message has been edited. Last edited by: electricpete,
You have very aptly described a lot about the nuke industry. In regard to AP 913, you are correct in assessing the litany of issues involved in the TOTAL flowchart. However, the very first block of the AP 913 flowchart, which is where everything begins, starts with identifying equipment important to safety and relaibility. That is RCM!
Unfortunately, the nuke industry does NOT have a corner on understanding RCM anymore than a fossil plant or a tire factory. Even with all of the bells and whistles and with the on-going internal and external assessments, with the dual verification of procedures, with the extensive tailboarding, etc, etc, the nuke industry does not have any more insight than any other type of plant or facility when it comes to identifying and understanding the precise nature of equipment reliability. It would take an extensive treatice to explain all of the ramifications of AP 913 and that is why I left my earlier response to Mac 13 for an off-line dialogue.
Being personally involved with the origin of AP 913, The Institute of Nuclear Power Operators (INPO), the Nuclear Energy Institute (NEI) and even the Nuclear Regulatory Commission (NRC) all struggle with exactly how to define equipment reliability and identify those equipment failure modes whose consequence of failure can have a unwanted effect on operations. As I have stated unequivocally in my book, it is not the KNOWN problem components that have the greatest potential to wreak havoc on one's plant or facility, it is the UNKNOWN consequences of failure. The known component problems are what maintenance budgets are made for. Components that even the janitor is aware will cause a problem when they fail is not what RCM is really all about. The heart of RCM is to ferret out the unknown, and innocuous component failure consequences, before they happen,....the one's whose failure will bring a plant to its knees should they fail under certain circumstances. The late John Moubray, one of my former colleagues whom I had known for over 20 years, was also in sync with this understanding.
Most major disasters that are not caused either by human error, acts of nature, or outright negligence, are, in fact, caused by equipment failures whose "consequence of failure" was previously unknown. Usually, after the unwanted event, the learning cycle of what caused that major event takes place. Unfortunately, that is a little bit late to find out what had previously been missed in the analysis. Sad, but true, the nuke industry is just as guilty of this indescretion as any other industry.
Author, "Reliability Centered Maintenance - Implementation Made Simple" published by McGraw-Hill
Thanks Gents for the replies
7 years ago ,We started RCM analysis at our Nuclear power plant ,4 years ago ,we are thinking about implemennt of INPO'S AP-913,as said by Mr.Neil Bloom ,the first parts of AP-913 (Scoping and Identification of Critical Components,Critical, Non-Critical, and Run-to-Failure)is one step of RCM or FMEA with the predefined "Critical criteria",so one of the output of RCM is the critical compontents list or data base,after the first step of AP-913,The fouth step of AP-913 (Continuing equipment reliability improvement),IMO, is a living RCM Procress to select the right task (predicitive and preventive)to manage the Failure consequence.
Neil, do you have examples for "unknown consequences which RCM will identify" which you mentioned in your above post?
Thank you for asking about the "unknown consequences" that RCM can identify. Your inquisitiveness in this regard is indicative of the level of understanding that is just beginning to take center stage in the world of RCM and reliability in general. Typical examples are prevalent throughout industry. That includes industries and facilities as diverse as a manufacturing plant, a chemical plant, a refinery, a commercial aircraft, the space shuttle, a nuclear power plant, an off-shore oil drilling platform, or even a shoe factory. The "unknown consequences" most often manifest themselves due to the lack of understanding of how to really analyze "hidden failures" and the inaccurate categorization of run-to-failure as a choice of strategy for a component that is NOT really run-to-failure. It is the "unknown consequences" that have the greatest potential to cause a catastrophic event.
As I mentioned in my previous post, the late John Moubray, who was a colleague of mine for over 20 years, was also in agreement with me on this issue.
A very typical example, that I mention in my book, occurred, not in a third world country but right here at a U.S. nuclear power plant. One of the major safety systems of the main 1150 megawatt turbine was considered so important that the designer included "triple redundancy" so that a runaway turbine due to an uncontrolled overspeed would never occur. The bottom line is that..... a runaway turbine, due to an uncontrolled overspeed, is exactly what happened! The turbine blew apart and only by luck, no one was killed. It caused several hundred million dollars in damages and resulted in the plant being shutdown for almost one full year resulting in an additional $600 million dollars in lost revenue. That catastrophe, which was totally avoidable with an understanding of some of the concepts I write about in my book, resulted in approximately $1,000,000,000 (yes that's one BILLION dollars) in total losses!
This total disaster could have been avoided with the addition of a few simple but strategic PM's that would not have cost more than a few hundred dollars to implement. The incorrect categorization of run-to-failure and the "unknown consequences" surrounding this event are all too common. Let me also set the stage here for something else to consider. A nuclear power plant probably has more technical expertise on site than any other entity. There are several hundred engineers, scientists, reliability experts, quality assurance inspectors, component experts, maintenance specialists, and a whole host of many other technical types that reside on site. The "unknown consequences" and the pre-existing conditions for the aforementioned disaster went under everyone's radar.
To provide you with another real-life recent example, consider the fire and explosion at the BP oil refinery in Texas. Seventeen people were killed and many others injured. This incident was so significant that BP convened a special investigation panel led by the former Secretary of State, James Baker. The Baker panel's final conclusion was that the explosion was PRIMARILY caused by an inadequate preventive maintenance program whereby certain component functions and failure consequences were apparently not well understood.
I could go on and on with a litany of other such typical examples.
The theme I am bringing to industry in my book and in my speaking engagements throughout the country is that it is not the "usual suspects" that cause the major disasters. It is not the components that everyone knows is a problem that have the greatest potential to wreak havoc on your facility; it is the "unknown failure consequences" and the incorrectly invoked categorization of run-to-failure that poses the greatest threat to safety and reliability.
Something you might find of interest, which is pertinent to this discussion, is an article I recently wrote for the October issue of Terry's Uptime Magazine published by Jeff Shuler. If you did not see the article, I have attached a copy of it below. It is titled "What is RCM Anyway?"
Author, "Reliability Centered Maintenance – Implementation Made Simple" published by McGraw-Hill
FROM THE OCTOBER ISSUE OF UPTIME MAGAZINE
WHAT EXACTLY IS RCM ANYWAY?
Reliability Centered Maintenance, or RCM as it is called, is a term used in the reliability community by different folks to mean different things. Reliability people in various industries around the world truly want to improve their preventive maintenance programs and there are innumerable of ways to accomplish that goal.
Unfortunately, and all to often, when the term RCM is used, it unknowingly becomes a convenient "handle" to add the aura of credibility or the image of some kind of technical authority to one's more simplistic approach to improving a preventive maintenance program. The thinking goes like this.... "After all, RCM was founded in the commercial aviation industry and if it is good enough to base the safety and reliability of aircraft on, it must be good enough for my plant or facility". While that's a true statement, many of us, unwittingly, I might add, try to fit RCM into their vernacular when, in reality RCM goes way beyond the simplistic vision of many preventive maintenance goals.
Through no fault of their own, most people do not know what RCM really is.
For case in point, let's look at some common thought processes that can be greatly misunderstood. For example, the phrase... "A strategic organizational realignment" might really mean ... "Let's fire the incompetents at the top and replace them with people who truly know how to run an organization." See how much better the former sounds? Or perhaps another typical example such as; ... "Notwithstanding a few minor technical impediments and cost challenges, we are very close to completing the project on a given timetable." In real parlance, this means... "We are way behind schedule and over budget." In the majority of instances, this is the same misunderstood syntax used when describing that "we are going to implement an RCM program to make our plant safer and more reliable".
From my experience with RCM and associated preventive maintenance program initiatives, approximately 80% of those people touting an RCM plan for which to improve their plant reliability, do not have a grasp of real-life RCM. In reality, what they actually want to implement is a PM Optimization program. A PM Optimization program is NOT an RCM program. What these folks actually want is to convert time directed overhauls into condition monitoring predictive maintenance tasks.... or they want to use cookie-cutter PM templates.....or they want to find better ways to schedule their existing PM's..... or they want to review the 20% of their known problem components that they believe cause 80% of all of their problems and reduce their costs correspondingly. Don't get me wrong here.... These are all wonderful things to do and I wholeheartedly endorse all of them as well as a whole host of other such peripheral betterment issues. But let me be very clear; these betterment issues are NOT RCM.
WHAT RCM IS NOT
RCM is NOT:
"¢ A remake of overhauls into condition monitoring.
"¢ Reviewing known problems.
"¢ A process that selectively picks and chooses a few given systems or certain
components to analyze, that everyone, including the janitor, knows is a problem and that has a major effect on the operation of the plant or facility when it fails.
"¢ Performing an analysis on a piece part such as a bearing or a shaft, for example.
"¢ Implementing a set of standardized task templates for PM activities.
The primary reason for implementing an RCM program is to identify components whose functional failures can cause unwanted consequences to ones plant or facility. If you already know what those components are you don't need an RCM program. You need a PM Optimization program which is a far cry from an RCM effort.
So why do people call these peripheral programs an RCM program? Usually, it is because it sounds better. It has a cache of professionalism, authenticity, and technical credentials associated with it. It is like buying a "NASA inspired space developed mattress" which sounds much more technically state of the art than describing it as a "foam mattress that does not move when you jump on it"!
IT'S THE UNEXPECTED
It has been proven over and over and over that the vast majority of major disasters which occur, that were not due to either nature, human error, or sheer negligence, were caused by equipment failures whose consequences of failure were unexpected and never analyzed, or those components which were incorrectly analyzed to be run-to-failure components. The disasters caused by these two reasons are "surprises" because they were totally unknown, unexpected and unanalyzed. Many of these disasters were caused by failures of rather innocuous or non-obvious components. A PM Optimization program would have little or no chance to ferret out those component failure consequences.
The "unexpected" disaster can also take place with the misguided conception that redundant components "automatically qualify for a run-to-failure" status. In the absence of identifying what I have termed "potentially critical" components, a hidden failure can go undetected if there is no indication of the failure and if there is no immediate consequence of the failure until another failure takes place in combination with the first failure.
After all, don't we all have maintenance budgets that encompass the known maintenance work that needs to be done? One rather ˜small' unexpected disaster can exceed that budget 2, 4, or 10 times over. Worse yet, one unexpected disaster can shut your facility down for good. The real-life examples of major disaster occurrences are bountiful with the latest BP explosion in Texas being just one of them.
WHY TAKE SHORTCUTS?
Even when we believe that we really do need an RCM program and not a PM Optimization effort, why do we take shortcuts that are commonly called streamlined or truncated RCM? We take them only because RCM has been made so difficult and costly to implement that it is mostly shied away from except for mega corporations with megabucks to spend. RCM usually ends up as an unsuccessful venture, even for the mega corporations. To put it in proper perspective, over 90% of all attempted RCM programs result in failure! This does not have to be the case.
Most people believe a comprehensive RCM program takes a team of 6 or more people, three or four years to complete. That could take at least 18 man-years! That's a scary thought and one that undoubtedly puts the kibosh on implementing a comprehensive RCM program. It does not have to be that way.
From personal experience in having developed and managed what is perhaps, even today, one of the most comprehensive classical RCM programs ever implemented, with over 125,000 components analyzed at a dual unit nuclear power plant, the "lessons I learned" are that for an average size facility, which comprises probably 95% of all facilities except perhaps, nuclear power plants and jet aircraft for example, a comprehensive RCM program for all plant components can be completed by a team of 3 to 4 technically qualified people in only four to six months!
PM BETTERMENT PROGRAMS
I totally concur with the late John Moubray, who was an acquaintance of mine for over 20 years, that there is nothing wrong with PM betterment programs; however, they should be distinguished from RCM. I am not bothered by the false name pretense of calling a PM betterment program an RCM program. I am concerned, however, about one's senior plant management being lulled into naively and falsely believing that his, or her, facility will now become more reliable and less prone to an unwanted disaster because the maintenance and engineering folks have "implemented an RCM program" borne directly from jumbo jets and nuclear plants when in reality all that is planned is a simple PM betterment program! In fact it is not fair to have senior management believe such a myth.
If you and your folks, as the responsible reliability liaison at your facility, do not know what happens with the unwanted failure of each and every FUNCTIONAL component in your plant, you do not have an RCM based reliability program. Note that I underlined, italicized, and bolded the word FUNCTIONAL. I did this because obviously there are components in every plant or facility that have no real function... they are there for convenience or for very minor importance. Think about it, if a truly functional component was designed into your plant, it is there for a reason. If its immediate failure has no unwanted consequence, and if a multitude of associated component failures in addition to the originally failed component have no unwanted consequences, and if it doesn't matter when the failed component is restored to an operable condition, then why is it in the plant to begin with?
SAE Document JA1011 has it place. It was not developed to be used on a pick and choose basis. In fact nowhere in the document does it say it can be used selectively. Remember, an RCM analysis is employed to "identify" components whose functional failures can result in unwanted plant consequences. If you are already clairvoyant and know what those components are there is really no need to pursue an RCM program. In such a case, RCM would be a waste of time and money.
Of course there are those components whose criticality is well known. That is a given. That is what maintenance budgets are made for. What about the other 80% of the plant that is not looked at? There are myriad critical components in that population just waiting to cause a disaster. In fact, it is within that "unanalyzed" population of components, that your disaster is most likely to occur! Keep in mind; it is the non-obvious and the rather innocuous components whose failure consequences have the greatest potential for wreaking havoc upon your facility.
BE AWARE OF THE RISKS
Let's look at the 80-20 program I mentioned earlier where only 20% of a plant is analyzed leaving the remaining 80% unanalyzed and probably in the run-to-failure category. As delineated in my book, I liken this to buying car insurance that insures you only while you are driving 65 mph on a freeway or while you are driving in heavy traffic, when you believe an accident is most likely to occur. You would not be insured driving on country roads, or driving slowly through your neighborhood to and from work, or driving on any nonbusy roads because it is assumed you would not have an accident under those conditions. You would assume the risk of having no insurance coverage during these times. Does that sound comforting? Statistically, when do most accidents occur? They occur within a few miles of home.
Many astute reliability professionals have begun to understand this logic. They understand that one unanticipated functional failure consequence can totally wipe out any routinely generated maintenance budget that was put together by including maintenance expenses only for the well known problem components that I refer to as the "usual suspects." They understand the risk of disaster that they assume by eliminating 80% of their plant from being analyzed for unwanted functional failures.
To put all of the aforementioned into a clearer and more focused picture, look at RCM as having three phases associated with it. The 1st phase is the heart and soul of the process. It is where the population of equipment requiring preventive maintenance is identified. This is the phase where RCM decision logic comes into play. Phase 1 is the "engine" of the process. The 2nd phase is to specify the tasks that will be scheduled on the population identified in phase 1. This second phase is where condition monitoring, predictive maintenance techniques and the use of cookie-cutter PM task templates are specified. The 3rd phase is the actual implementation of the specified tasks. This is where EAM and CMMS systems come into play.
Look at the 3 phases of RCM like you would look at a car. For example, the heart and soul of the car is the engine. Even though a car has many different facets to it such as tires, brakes, windshield wipers, seats, windows, and so on, it is the engine that singularly defines the car. All other facets of the car are peripheral to the engine.
In summary, don't unintentionally place your senior management into the false belief that you are going to make the plant safer and more reliable. If it is really a PM betterment program you are after, tell your management that you are embarking on a program to reduce known costs. There is a major difference between a program such as RCM which is designed for truly enhancing safety and reliability and its concomitant cost avoidance benefits (such as avoiding potential disasters) and a PM betterment program which is strictly an economic exercise implemented solely to reduce known costs.
In virtually every case, any type of true RCM effort will result in certain INCREASED costs because certain previously unknown failure consequences will be addressed and hence, those components will need to be continuously maintained within the preventive maintenance program. The real benefit of RCM is its ability to ferret out those components whose failure consequences were previously unknown so that they can be appropriately addressed to avoid an unwanted disaster. Obviously, avoiding unwanted disasters has enormous cost avoidance benefits. However, if your facility is such that the worst thing that can happen is within the realm of acceptance by your senior management, then RCM is not really needed. A PM betterment program is what should be pursued.
Neil Bloom is the author of "Reliability Centered Maintenance- Implementation Made Simple" published by McGraw-Hill. He is a mechanical engineer with over 35 years of both hands-on and senior level managerial engineering and maintenance experience in RCM and Preventive Maintenance Programs in the commercial aviation and commercial nuclear power industries. He is an international guest speaker on RCM and an instructor of RCM in the Continuing Education Division at the University of California – Irvine (UCI). Neil provides 3-day RCM Training Seminars and can be reached at neilbloom@RCMauthor.com or (949) 218-1286. His website is www.RCMauthor.com.
I enjoy reading your writing and experience. I've been letting this thread soak in and have formed an opinion I'd like to share.
I see RCM discussed sometimes as sort of an end all be all. I would define RCM as man's reasonable attempt to anticipate failure, attempt at failure avoidance, and to prepare for eventual failures.
I believe that man is not capable of full comprehension of, and anticipation of, failures in extremely complex interrelated systems. Luckily, most systems aren't that complicated. But there's always the incidence of coincidence, as I like to call it, that we humans fail to anticipate. Coincidence is a phenomenon that amazes and amuses me.
Consider the Biosphere 2 experiment. I'm sure that the scientists and engineers involved in that project attempted to anticipate every possible failure before the experiment started. What failed? I understand they failed to anticipate the extent to which higher life forms depend on microbes to balance nature and to continue life on earth. Sometimes it's the smallest things.
The space shuttle accident was an example of an unanticipated failure.
No matter how much we try to understand a system, there may be an interrelated failure that we failed to anticipate. I have not been involved in a nuclear facility level RCM effort. I'm sure I would be impressed, as I am with your approach to RCM. Most of the failures listed in our FMEA exercises are based upon our experience and imagination of the participants as to what could fail. I guess that's as good as it gets. You cover this in your article under "It's the unexpected". I guess the longer a site has been operating, the more real failures they have experienced and have prepared for. An RCM effort may shed less additional failures on the long operating plant and may be seen as less value overall.
In your article you separate the RCM analysis from the RCM group. The items you list in the article under "What RCM is not", don't most places use the RCM group to drive the items in the list though? If they have an RCM group, that is. In practice isn't an RCM group's function broader in scope than only performing RCM analyses? At our site our group takes on many projects such as spares analysis, CMMS data integrity (this is a huge effort), problem code analysis, PM optimization, assembling equipment manuals, and acts as liaison between equipment, crafts and operations. No other group has the natural tendencies that it takes to enjoy working on these projects. Natural tendencies also are why we have the "us and them" separation between functional groups. Human nature I suppose.
I'm glad we have this forum to further discuss this subject. I know I have a lot to learn. As a group we will move forward and get better recognition in industry.
You have articulated some very insightful comments and you seem to be pretty much in tune with some of the key aspects of Terry O's Forum. However, there is one issue, which, based on your comments, I need to further clarify.
I want to be very clear that the RCM analysis IS NOT SEPARATED FROM THE RCM GROUP! The RCM "team" is responsible for all facets of the program including spares analysis, CMMS data integrity, problem code analysis, PM optimization, equipment manuals and all of the other items you mentioned. I want to make it very clear what RCM is and what it is not. As I mentioned in my previous post, IN AND OF ITSELF...
RCM IS NOT:
"¢ A remake of overhauls into condition monitoring PdM tasks.
"¢ Reviewing known problem components.
"¢ A process that selectively picks and chooses a few given systems or certain components to analyze, that everyone, including the janitor, knows is a problem and that has a major effect on the operation of the plant of facility when it fails.
"¢ Performing an analysis on a piece part such as a bearing or a shaft, for example.
"¢ Implementing a set of standardized task templates for PM activities.
The point I am making is that all too often, folks will pick out any one of the above, and call that individual task an RCM program. Any of the above, IN AND OF ITSELF, does NOT constitute an RCM program.
As I also mentioned in my previous post, RCM consists of the following three phases:
PHASE 1: This is the Phase where the components that are required to be in the
preventive maintenance program are identified. This is where the RCM
decision logic and the RCM analysis is invoked.
PHASE 2: This is the Phase where the various PM tasks, including condition monitoring
PdM tasks, are specified for the population of equipment identified in
PHASE 3: This is the Phase where the PM (PdM) tasks specified in Phase 2 are
executed. This is where EAM and CMMS systems come into play.
So RCM includes ALL of the three aforementioned Phases, with Phase 1 being the engine driving Phases 2 and 3. The clarification I want to make is that all too often, folks will only look at Phase 2, IN AND OF ITSELF, and refer to that effort as an RCM program. That is an incorrect consideration of what RCM is. Just changing time directed overhauls into condition monitoring PdM tasks, IN AND OF ITSELF, is NOT RCM. Likewise, using EPRI cookie cutter PM templates, or any PM templates for that matter,
IN AND OF ITSELF, is NOT RCM.
Similarly, ascertaining the most effective EAM or CMMS system to use, IN AND OF ITSELF, does NOT constitute an RCM program.
So, yes, RCM includes ALL of the three Phases mentioned above and the RCM group is involved in ALL aspects of the preventive maintenance program..... that includes Phase 1, Phase 2, and Phase 3.
In summary, an RCM analysis of a nuclear power plant is no different from an RCM analysis of any facility. As I mention in my book,....
"RCM is applicable to any plant or facility where it is unacceptable to incur an unplanned shutdown, a loss of production or generation capability, a regulatory violation, environmental hazards, explosions or personnel injuries." "In essence, RCM is applicable to any entity that manufactures a product or produces an output where it is unacceptable to incur unplanned interruptions of the operation, or worse yet, an unwanted disaster!"
The three Phases I noted above, are applied the same to any type of entity.
Author, "Reliability Centered Maintenance – Implementation Made Simple" published by McGraw-Hill
Neil, can you fit the Moubray's RCM 7 questions into your 3 phase RCM approach for clarify purposes?
Let me try to summarize what I understand from your long reponse above:
1) Basically, you say RCM should be comprehensive and not "streamlined or trunscated".
2) You also say RCM will help to find more hidden failure or "unknown failure consequences".
I really like to see the power of RCM to identify unknown failure consequences. Can you provide a list of more examples from your experience to strengthen this point of yours further?
More equestions to further understand your view about RCM and reliability:
How often to test trip and control instrument loops?
How often to test pressure safety valves (PSV)? Is it important to pre-pop PSV before overhauling them?
How to operate 2 parallel pumps? Do you prefer pump switching or duty/standby concept?
Lastly but not the least, is it possible to post a picture of you and your 20-year aquiantance, the late Mr. John Moubray in the Photo Alley of this forum?This message has been edited. Last edited by: Josh,
Thanks for your questions. First, let me clarify that the seven questions were not John Moubray's seven questions. They are the seven questions of SAE Document JA1011. In regard to your question as to whether the 3 phases of RCM that I mention embrace those seven questions, the answer is emphatically, Yes!
In fact, those seven questions, in addition to the other specificity contained in JA1011, such as the distinction of separating strictly economic consequences from safety or operational consequences and even going further by separating safety consequences from operational consequences were all included in the RCM program I developed in early 1991 which was long before they finally came out in the SAE Document in 1999. Additionally, I was also very specific in segregating hidden failures from evident failures which is now a major part of the SAE Document. I explain all of this in detail in my book.
Now let's look at your next question.... Yes, RCM should be comprehensive, not streamlined, truncated, etc. I am not saying these other shortcuts cannot be used. What I am saying, however, is that they are not sufficiently robust and will have very little likelihood to prevent the types of disasters, catastrophes, and other major unwanted events that could occur. Like myself, John Moubray was very much in disagreement with shortcut versions of the process. More importantly, did you ever wonder why these streamlined, truncated, and other shortcut versions ever came into being? It was primarily because of the difficulty of implementing a comprehensive RCM program. Bona fide RCM has become so convoluted and obfuscated over the years that it has virtually ceased to become the useful tool that it was always intended to be in its origin in commercial aviation. Why do I know this? Because I spent many years, both as a practitioner and as a member of senior management, in the engineering and maintenance organizations within the commercial aviation industry where I worked with MSG-2 and MSG-3 Logic which were the forerunners to RCM. RCM was never intended to be so overly convoluted. When RCM left the commercial aviation industry, it morphed and transformed into something so complicated that it became known as..... "a vast wasted money pit", "a process not worth doing", "a total waste of time", "an accumulation of stacks of useless analyses laying on shelves all over the place", etc, etc. Unfortunately, and unjustly, and far too often... that perception still exists today.
For your next question regarding hidden failures.... It is my opinion, and I have written about it extensively in my book as well as numerous technical publications, including Terry's Uptime magazine, that addressing "hidden failures" is perhaps the greatest protection one can have to ensure the safety and reliability of his, or her, plant or facility. The "known" problems are just that...they are "known". That's what maintenance budgets are for. There is no real challenge in addressing known problems other than specifying the most efficient type of PM task whether it be an overhaul, replacement, or some other type of PdM task I devote an entire chapter of my book to the understanding of "hidden" failures and their consequences. Understanding "hidden" failures and understanding when to correctly invoke a "Run-to-failure" strategy are the fundamental keys to reliability success.
In regard to testing instrument control loops.... Unless specified by a regulator such as the FAA, the NRC, the EPA, etc, the decision is basically determined by prudent engineering judgment also taking into consideration, vendor and manufacturer recommendations. Note that I did not say "exclusively" using the vendor or manufacturers recommendations because oftentimes those recommendations are far too conservative. There are innumerable "inputs" for selecting the optimum interval for any PM activity and I address them all in my book.
In regard to PSV's..... PSV's fall into several categories. There are some that have state, and/or other federal requirements to be tested at certain intervals. These are most often associated with pressure vessels. For the vast majority of the other PSV's, the testing requirements are once again established by prudent engineering judgment. A very novel approach that I used was to specifically categorize those PSV's that were not governed by state or federal law, and those whose failure consequence could not cause personnel or safety hazards....and establish a "random sampling" program to ensure that by testing the correct sample population, there was a 99% confidence level that the balance of that population was acceptable. This program required very precise instructions and details and also required the commensurate approvals of the senior hierarchy of the technical staff.
Should PSV's be pop-tested before overhaul? In my opinion, yes, they should. The reason is strictly for data collection to ascertain the characteristics of the PSV such as... Was it in tolerance? How far out of tolerance was it? How long was it in service? What part failed? Could its environment have influenced its pop test results? We found that by capturing the pop-test results and statistically analyzing them, we could more precisely learn about make/model disparities and more accurately determine optimum replacement intervals.
In regard to operating two parallel pumps..... This is the age old question and it really depends on you the operator of your plant. There are advantages and disadvantages to both. There are some who run only one pump until "wear out" occurs and then transfer to the other standby pump. The other standby pump, however, must be maintained with a robust PM program including intermittent operation to ensure it will operate and function when called upon to do so. Some people prefer to rotate the use of parallel pumps. This tends to extend the total operating time between replacements. This practice will, however, wear both pumps equally so that when one fails, the other is probably not too far behind. Obviously, this can result in several problems. There is always the question of spares. In the first case, perhaps only one spare pump may be sufficient. In the second case, it would be prudent to have two spares available. In either scenario, Corrective Maintenance must be performed on the failed pump in a timely manner. Though in the latter scenario, Corrective Maintenance on the failed pump has a much more critical time frame.
As for pictures.... I do not have a 20 year old picture of myself nor do I have one of John Moubray. However, my website does have a rather recent picture of me. You can access my website at: www.rcmtrainingseminars.com.
Author, "Reliability Centered Maintenance – Implementation Made Simple" published by McGraw-Hill. ISBN 0-07-146069-1
In the 4th paragraph, there should be a period after the phrase... "...some other type of PdM task.
Thanks for your reply, Neil.
Actually, I asked you to post a picture of yours (at any age) together with John Moubray who was your 20-year acquantance. Anyway, you look like Albeit Einstein in your website (with your hairstyle in the picture).This message has been edited. Last edited by: Josh,
|Powered by Social Strata|