Does anyone have experience applying RCM to facilities systems (HVAC, chillers, boilers, etc.)? I am curious about issues that may be specific to this type of equipment, e.g. determining system boundaries, dealing with lead/lag arrangements, systems that serve multiple buildings. In the past I applied RCM to manufacturing systems where failure consequeneces were more obvious and costs were easier to calculate. I would be interested in any advice from someone who has worked with facilities equipment in a government or laboratory environment.

Thanks!

Shelley:

Yes, US Coast Guard, US Navy and General Motors Facilities here.

What would you like?

I am also responding to your email about the NAVSEA RCM training next week. I am just awaiting info from the program director.

Howard

Right now I have a few questions.
1. We have a lot of equipment that is needed only intermittantly, but that would be considered critical when called upon. Do you treat these as hidden failures?

2. Many systems have lead/lag equipment. I'm not sure what the schedules or procedures are for these to be rotated. What key questions do I need to ask? Also, just following the RCM logic in my head, if there is redundancy it may be that the majority of the analysis would end up as "no scheduled maintenance," and somehow that doesn't seem right. Do I need to make adjustments to the process when analyzing these types of systems?

3. Because of the situations described in questions 1 and 2, many of our electrical panels and MCCs are not sufficiently loaded most of the time to produce a meaningful IR thermography report. We have a well-established IR program (electrical only), but I am concerned that we are not getting value out of their efforts. Most of their reports indicate that the panel was insufficiently loaded when the inspection was done.

I am definitely still a novice in this arena, but I have responsibility for getting RCM established and integrating the PdM and PM programs. These significant differences from a manufacturing environment have me stumped for the moment. Any help will be much appreciated!

quote:

1. We have a lot of equipment that is needed only intermittantly, but that would be considered critical when called upon. Do you treat these as hidden failures?

Yes. The first question in most logic trees will usually ask if the failure is obvious to the operator. If the answer is 'no,' then it is a hidden failure. There will be exceptions, based upon how you set your boundaries. For instance, if you set your boundaries around just the equipment, then it may be obvious to the operator that the equipment malfunctions. However, if you include the effected components, then it may not be obvious until it is required, which would be a yes, it is a hidden failure.

quote:

2. Many systems have lead/lag equipment. I'm not sure what the schedules or procedures are for these to be rotated. What key questions do I need to ask? Also, just following the RCM logic in my head, if there is redundancy it may be that the majority of the analysis would end up as "no scheduled maintenance," and somehow that doesn't seem right. Do I need to make adjustments to the process when analyzing these types of systems?

No, not really, other than you may wish to ensure that there is lubrication and servicing performed, where required, especially if equipment is being cycled.

quote:

3. Because of the situations described in questions 1 and 2, many of our electrical panels and MCCs are not sufficiently loaded most of the time to produce a meaningful IR thermography report. We have a well-established IR program (electrical only), but I am concerned that we are not getting value out of their efforts. Most of their reports indicate that the panel was insufficiently loaded when the inspection was done.

It sounds as if infrared is not truely effective. Perhaps motor diagnostics, resistance tests or voltage drop tests would be more effective.

See my article, "Your Maintenance Program Is Effective... Isn't It?" in the first edition of Uptime Magazine on p. 32. If you have not received a copy of Uptime magazine, contact ReliabilityWeb (tohanlon@reliabilityweb.com).

Sincerely,
Howard

implementation of RCM

hello..i'm a new member here and currently doing undergraduate programme in mechanical eng.
my interest is about gas compression system. the thesis is as the topic stated.
would be appreciated if anybody like to share their experience/suggestion or ideas. thank you

Shelley / Howard,

I can't agree with the statements above, however it may be that I dont have all of the information. Particularly this statement:

quote:

1. We have a lot of equipment that is needed only intermittantly, but that would be considered critical when called upon. Do you treat these as hidden failures?

Answer from Howard:

quote:

Yes. The first question in most logic trees will usually ask if the failure is obvious to the operator. If the answer is 'no,' then it is a hidden failure.

This is not correct Howard. The actual question at the beginning fo the RCM decision diagram is "Is the loss of this function evident to operators under normal circumstances". (SAE JA1012, Diagram 2, page 17 (I think its that page)

It is a common error to confuse visible with evident, but if it will ultimately become evident to the operator then it is not hidden at all.

In the case of intermitently used equipment, as soon as this item is used it will become immediately evident that there has been a failure, so it is evident. (Not Hidden)

To which the statement is generally, but then its too late. Yup! Thats why you need to put some form of management regime in place to manage this.

A hidden function is one that requires something else to occur prior to becomiong evident. Within the standard, or the guide to the standard I don't recall exactly, it states that "Hidden functions, or failures, by themselves have no consequences".

They only become evident once there has been another failure. In the case of intermitently used equipment, provided it is normal operating circumstances to use this equipment from time to time, then there is no additional failure. merely the requirement to use the equipment, not abnormal operating circumstances.

If the intermitently used equipment is either standby equipment or is a protective system such as fire water systems, then things change again. There would need to be an additional failure (of the duty pump, or a fire) so therefore the item would be hidden rather than evident.

There are dramatic reasons behind this that have to do with the way that the decision diagram, or the RCM decision logic, manages failure modes. That is, if the failure mode by itself has direct consequences, (that is without requiring another failure mode) then there is a need to take some form of preventive, predictive or (if warranted) modifying action.

If there are no consequences by themselves, that is there is a need for another failed state or failure of some form, then we are trying to manage the probability of a multiple failure.

This is a pretty stark issue within RCM and one that at times takes some time to come to grips with. There is the common tendancy to confuse not visible, or not visible for a time, or safety consequences, with hidden. But there are very strong reasons why they are separated.

RCM and cycled equipment

The RCM logic will assit you in determing the correct frequency for rotating duty cycles, through cost optimization equations normally, it will also help you to question traditional measures such as lubrication etc.

The statement that if there is a stand by then everything becomes run to failure is not really correct Shelly. Throughout the RCM logic there is a prompt for you to look at the cost of performing the maintenance versus the cost of not performing the maintenance over time. (Have a look at SAE JA1012 diagrams 1 and 2 on pages 16 and 17)

So were there are dramatic problems with high costs of failure, high secondary damage, significantly shortened life etc, then there may be a case for traditional time based maitnenance regimes. (But there also may not) At all times the focus is on finding the most cost effective maitennace regime for failures with economic or operational impacts.

When we are talking about safety or environmental consequences it becomes an entirely different story altogether.

One thing that you need to check, I would suggest, is why they are being rotated in the first place. During the past twenty years there has been a sort of group think regarding the need to rotate motors and pumps 50% each. This is almost absolute foolishness in just about all cases. In items where there are wear related failures, or time based failures such as pump impellars, all you are doing is ensuring that bo0th will fail at similar times.

I hope this is of some use.

Daryl:

Please read the full explanation prior to commenting, or, if quoting, include the complete statement and not just the first sentence. We had this problem in the past.

I don't see anywhere where there is disagreement.

However, you do mention that is not a good idea to run redundant equipment. Equipment such as electric motors and pumps will degrade over time by NOT operating them. They do require some level of operation to ensure that bearings are lubricated, the insulation system is exercised, material is not collecting in the impellor and the seals are exercised.

If the components of a pump failed/wore at the same rate, then what would be the point of tools such as RCM? or, Wouldn't that set the condition that you could accurately template RCM?

Howard

Hoard,

We have had this problem before Howard, please read the entire posting before commenting.

This is the full quotation of what you wrote, and regardless of how much of it I put here it remains at odds with what I wrote.

quote:

There will be exceptions, based upon how you set your boundaries. For instance, if you set your boundaries around just the equipment, then it may be obvious to the operator that the equipment malfunctions. However, if you include the effected components, then it may not be obvious until it is required, which would be a yes, it is a hidden failure.

The fact that it is not obvious until required does not make them hidden as stated in the previous posting. If this were the case then they would be evident, if it was not obvious until there was another failure, thus making them needed, then this changes everything.

As for the second point you are either deliberately or mistakenly misunderstanding the point I was making.

I never said dont rotate them, I argued the point that 50/50 rotation is one of the hieghts of stupidity in most cases. This is quite a different thing.

You can accurately template RCM. But there needs to be a fair amount of control over the process so as to make it realistic.
(But thats not in dispute here)

Shelley,

Time of detection is not an issue in deciding whether the function is evident or hidden to the operator in the normal course of his duties. What matters is whether the consequence occurs at the same time as the failure, irrespective of when the operator found this out. If it needs a second event, such as a gas release or running pump failure before the operator knows the consequence, then it is a hidden function. Just ask the question - does the consequence occur at the same time as the failure, and the answer should tell you if it is evident or hidden.

V.Narayan.

Thank you, Vee. My point, exactly.

Unfortunately, the need for 'one-upmanship' seems to continuously cloud responses.

Howard

Howard,

This is not the case. I read something that appeared to be incorrect and, as I think this is avitally important issue, I corrected it.

I didn't take this to be your point from what yoiu wrote.

Gentlemen,
Thank you all for your insight on this topic. I believe my original question about hidden failures was somewhat vague, which may have contributed to some misunderstanding. However, I have benefitted from each posting and can now approach the hidden failure question in the decision diagram with more breadth of knowledge. Vee, your single statement is definitely a good litmus test.

I have another question. In the past I have simply used my own Access database to input, track, and report RCM analysis information and findings. I'm curious if there is an advantage to using commercial software for this, and if so where can I look to find some? Our CMMS is Maximo, so it would be nice if the tasks could automatically dump into Maximo. Perhaps I should start another thread for this one.

Thanks,
Shelley

Shelley,
There are many advantages to commercial Equipment Reliability software, but I don’t want to turn this in to a marketing thread.

PREMO XPERTS, our reliability software, is fully integrated with Maximo. There are DOE and other US government agencies using our software and services.

Please contact me at LarryJohson@fractalsolutions.com for more information.
Regards,

Interested to know what can that PREMO XPERTS do for reliability?

Shelley, You may find some advantages in using commercial RCM software, I would suggest as I have in other areas of this forum that you look at the cyberalfa package by DEI of Aberdeen. Their web address is http://www.dei-ltd.co.uk This system outputs a data base that is compatable with most CMMS / EAM systems, or at least could be mapped to fit the system you have, You will also find the inventory management bit of the software useful. Please let me know what you think of it.
Should you need further information or assistance please let me know.

Shelley,

I led an RCM project at the Savannah River Site (DOE) many years ago (10+), and I am occassionally in contact with the guy who took over that project. If you haven't been in touch with them, you might want to compare notes and practices. I can get you an e-mail address if you are interested.

alhr@novonordisk.com

AlanH2,
Thanks. I did speak with SRS a couple of months ago regarding their RCM and PM Optimization program. I've also learned a lot from AEDC at Arnold AFB. Lots of good experience out there, and forums like this one have been extremely helpful to me. Thanks for responding!

Shelley

Shelley

I hope your RCM project is progressing at the pace you predicted and projected.

I thought you might like to see this paper from RCM-2006:

Reliability Centered Maintenance (RCM) in the Facilities Environment – New and Existing Buildings by Alan K. Pride, Associate Director, Smithsonian Institute

Please hold April 3-6 open for RCM-2007 in Honolulu!

Aloha

Terry O

Terry,
Thanks for the paper! I have ordered the CD of the proceedings from RCM 2006. There were quite a few presentations that I look forward to reading. As for next year, Honolulu is a tough sell to management, so we'll see.

Our first RCM analysis went very well, and we are now in the process of training more people and scheduling future projects. I have been encouraged by the level of acceptance and participation at all levels of the organization. A big learning was that operator/craftsman participation throughout the process is absolutely essential. Culture change is slow and difficult, but I believe RCM is the best tool to get us there.

Thanks for this website and the conferences that help all of us continue to learn!

Shelley