From the RCM II bible page 70

When listing failure modes, do not try to list every single failure possibility regardless of it's liklihood.

Only failure modes which might reasonably be expected to occur in the context of the question should be recorded. A list of reasonably likly failure modes should include:

- failures which have occured before
- failure modes which are already the subject of proactive maintenance routines
- any other failure mode which have not yet occured but which are considered to be real possibilities

Steve:

quote:

I dont think it matters about the detectability of the maintainer to the spefic mechanism of failure. At the FM stage of the analysis we should not be making assumptions about detectability. My view is that analysts should be writing what fails and how. The right team would know what the resuling failure types might be should the component be stripped and a detailed examination done.

My point for consideration: If a complex item fails and it is not reasonable that the maintainer/ operator could decern why it failed, it is important to describe HOW it failed and describe the WHY in a consistant (for all analysis you do) manner. It is common practice for a failed pump to write "item fails no output due to internal failure". Then if warrented an additional analysis would be done on the LRU (Line Replaceable Unit) considering it on a teardown bench where the bearings, gears, seals, pistons, ect would be analysied.

quote:

(facilitators hat)I would not list incorect installation - since we would never end the analysis of the plant.
I would write lack of lubrication - because that ends up as a lube task.

Steve - I agree that you would not apply "incorrect installation" as a blanket FM for every component - only for those you know are caused by this.

Johnny is correct - the right team is important and finding where the problems lie and using this information to construct your FMs. I worry that we haven't used Steve enough and actually based our analysis on the to and fro of questions to him. My fault.

Mike.

Hello Forum,
I maybe jumping ahead with the direction of this topic but would like to show a fin fan strategy I was involved with developing a number of years ago on a petrochemical facility. We had some serious reliability problems on our fin fans at the time. It appears very similar to Steve’s example.
We gathered a team, performed a FMEA, developed the strategy and implemented it.
It was developed using RCMCost.

Please use the FMEA attached if it helps in your online discussions.

Cheers - Gary

Fin_Fan_Heat_Exchanger_FMEA_V2.xls (22 Kb, 111 downloads)

Gary,
How did you come up with the intervals?
Can you explain the method please.
Rgds
Steve

Gary,
also, did you manage to get the before and after results?

Gary,
Very good FMECA. This example follows the guidance on functions discussed eailer.
Same question on intervals.
Gary, have you used other software other than RCMcost, what capabilities do you like with RCMcost? I have read up on some of the functionality of the software with great interist but have not used it. Contact me off line or in another thread if you would.

Thank you for the example provided. By all means, do finish the story of implimentation and benifit/ RoI.

V/R:

Group

You are getting some attention in the RCM world and we have just received an offer for the group to use an online web based version of RCM Worksaver to make this project flow easier (possibly).

If you are interested please let me know and I will arrange.

Terry O

Terry,
I think it is best to avoid using software products. There are many on the market including ours that could be used. The group has decided to use an RCM2 approach and we are happy to form other groups to use other methods. Gary has posted the results of an RCM Cost analysis. It will be interesting to compare the results at the end of the program.
Thanks to RCM Worksave for the offer, but unless others in the group are highly in favour, I would prefer to use a spreadsheet.
Rgds
Steve
www.reliabilityassurance.com

I have the same feeling as Steve about the software, the software could be seen as a silver bullet while it can be done with excel and a dose of horse sense.

About the intervals on Gary's case:
It makes perfectly sense, most pm jobs starts at 3 months and multiples of that. The trick is to combine all the necessary activities in one event that can be planned in advance.

Hello Steve,

quote:

“How did you come up with the intervals?”
The lubrication intervals were developed with an internal lubrication specialist.
The CM intervals were developed with a combination of CM specialist advice and P-F intervals.
The remaining strategy was developed using failure history, weibull analysis, local knowledge and manufacturer’s recommendations.

quote:

“Can you explain the method please.”
We developed a cross functional team from site with previous knowledge of the equipment (the fin fan was a part of a larger site FMEA). I had already done a pre-model build in RCMCost that we validated during the FMEA session. The strategy was then implemented and monitored.

quote:

“Did you manage to get the before and after results?”
I did manage the before and after results based on a number of maintenance KPI’s - MTBF, MTTR, FBPM and Availability. I should be able to get the results from an ex colleague if so required but don’t believe that’s the purpose of this reliability thread. If I can get them, I don’t mind sharing them.

quote:

“Gary has posted the results of an RCM Cost analysis. It will be interesting to compare the results at the end of the program.”
I don’t mind you comparing the ‘fictional’ results with my ‘actual’ results at all. Providing we all remember that my results were real and based on real data.

The purpose of me sharing my FMEA was too add some weight to some of the interesting debate seen so far. I can only share with you what has and hasn’t worked for me. So far the RCMCost methodology has delivered. I would however agree with Steve and continue your online study using the spreadsheet approach. I’ll log back into the forum in March next year to see if you have managed to get any results (tongue in cheek!!)

Cheers - Gary

Hello Johnny,

Thank you for your compliment on my FMEA.

quote:

Gary, have you used other software other than RCMcost, what capabilities do you like with RCMcost?

I have not used any other RCM software other than RCMCost. I used it as an end user for 3 years and now as a consultant for 4 years. My FMEA example was from my time as an end user. I have used a spreadsheet approach in a previous role, but not for at least 7 years!It would be unfair to talk about the software capabilities in this reliability thread, if you do require more info, I am willing to contact you offline.

Cheers - Gary

Hi, gentlemen. I´ve been following the thread in silence from the beginning. I totally agree with Rogier regarding his statement from the Bible:

"Only failure modes which might reasonably be expected to occur in the context of the question should be recorded. A list of reasonably likly failure modes should include:

- failures which have occured before..."

Now, if installation is an issue in this place, I suggest including "...due to bad performed mounting" as a FM. It may follow that training is a proposed task, which otherwise wouldn´t be evident.
Open to discussion...
Regards

I wouldn't include

quote:

..due to bad performed mounting

, because that brings us to Root Cause Analysis

If we know up front that the craft is not prepared to handle the repairs, and we do nothing to correct that, it is a failure of management, not the equipment...

quote:

"Only failure modes which might reasonably be expected to occur in the context of the question should be recorded. A list of reasonably likely failure modes should include:

- failures which have occurred before..."

It is critical to consider more than failures that have occurred before. Most critical (the kind that kill people) failures are "almost" completely designed out. We need to ensure we consider those things that almost should not happen. This is even more important when you are doing any type of maintenance optimization. Existing inspections may seem useless if you are looking for things that have happened.
I don't think this is expressed well, but it's early and I think everyone can get the point.

Rogier,
Facilitator's job is to keep the group moving . Classic cause of RCM failure is no action.
I think we should be writing Functional Failues and Failure Modes.
Rgds
Steve

Steve, good call. Who has the dot? Let’s try to get it moving again.
On Monday, I am ready to go. lol, everyone needs a vacation once in a while. To all celebrating a holiday this week, safe festivities to you and yours, to anyone else, same wishes for the weekend.

Hi Steve,

Thanks for the reminder .
Perfect input so far and good discussions.
I will have made a update from the results before end of the week.

Refering to Terry's question regarding the RCM Worksaver I also agree to proceed in the way we are doing. RCM is thoughtware not software.
It's nice we have the attention of the RCM world. When this study is finished it's nice to make a new topic with a overview of RCM software and compare the pros and cons them.

Keep up the good work.

Regards,

Rogier

All,

New versions ready for comments and updating.

Ground rules defining FM defined in Projecplan V4
If how/why not specified it's assumed as "for any reason"

Workout of recurring FM there where it is assumed to happen first
Projectplan version 4

RCM2 template V3

For a better readability I attached information worksheet also in pdf.

Regards,

Rogier

rcm2_template_v3.pdf (33 Kb, 58 downloads)

Here is the downloadable excel file V3.

rcm2_template_V3.xls (78 Kb, 68 downloads)