Ladies / Gents,
If i can get some help on the following subject it would be truly appreciated.
I'm hunting for a risk matrix / 20 questions / consequense probality technique / system that i can use to determine, according to the priority ranking, for which equipment i need to develop strategies first. It needs to take everything around equipment into account, failure, maintenance, process impact, cost, spares etc...

kencj

Kencj

I have an answer – I will not say if it’s the best or worst – you can decide.

I suggest using the attached spreadsheet as a starting point. Its purpose is to filter equipment based on failure consequence. Partnered with historical costs/losses (which should cover the more probable cases) one may understand what equipment is the highest priority for further analysis using RCM/FMECA.

For spare parts and such I do not have an easy answer.
First, One must uncover which equipment has the highest consequences of failure or is the most probable (again historical costs should help identify the most probable).
Second, One must analyze this equipment with an RCM type process.
Third, Once a maintenance strategy is known (developed in the RCM process) then one can look at spare parts.

For example: If the decision has been made to perform condition based monitoring then the warning time to imminent failure versus the lead time of parts will be the basis for decision making. If RCM is not performed it is a pretty nebulous decision.

Please see the first three posts in “Reliability and Availability Levers” which is inside of the Forum “Posts about Improving Reliability”.

shared.xls (16 Kb, 83 downloads) Functional Failure Consequence Matrix

Actually, I wonder if this thread should be in the Forum "Posts about Improving Reliability".

Thanks Jaz,
I was wondering about the posting as well and wanted to change it but for some or other reason i was unable to move it.
I concur, it should be in another posting list.

Kencj

Kencj,

Does this information help? Do you have any further questions about this information?

Tx Jaz,
I was able to build something around the information you gave me and also came across some software called MPDS which also adresses all the things in their risk matrix that i think one should consider.

Kencj

I wasn't able to locate MPDS software. Can you provide a link?

Thanks

quote:

Originally posted by Joe Petersen:
I wasn't able to locate MPDS software. Can you provide a link?

Thanks

Joe,

MPDS is Strategic's old RCM software (ca. 1995). I think they changed its name or they no longer support it.

R,

Larry,

MPDS was actually BHP Engineering's software. I was told in 1994 by one of the product originators, that it was based then on Tony Kelly's "Strategy Based Maintenance" and I believe it is still in use in a number of BHP companies and by many consultants who in the past were involved in developing the system. (Most maintenance engineers in Australia seem to have had a go at it at one time or another)

When BHP engineering was sold off to Hatch Engineering out of Canada, MPDS and a range of other products were sold to Strategic were they were rebranded and rebuilt in some cases. Other products from this period include "Lumpies" now called "Ca$H", and "SCAS" (spares criticality and assesment system)

MPDS was the basis for what is now called RCMTurbo.

Best regards,

Dear kencj,

I would actually not recommend starting an analysis with a criticality type of approach such as Fuzzy logic matrices or other quantifiable criticality matrices.

The reason for this is because the only way you can really determine criticality is once you understand the ways it can fail and the effects of each of those failure modes.

So, if you want to make any form of informed decision based on criticality you first need to go through the arduous task of defining the functions, failures, effects of the failures, then quantifying these on your matrix.

What I suggest you do is look at a prioritization approach, rather than a criticality approach.

THis sounds like semantics at first but it really is a rapid way for you to get moving with your analytical work while ensuring that you are dedicating your efforts to the assets that have the most impact, as agreed by your companies decision makers.

The best multi-criteria decision making methodology available, and one that I use regularly, is called Analytical Hierarchical process.

It is rapid, thorough, and will give you an answer that will determine the best way forward based on current corporate objectives.

A quicker way to start your process I would suggest. Criticality can be graded as you are mving through your reliability analyses.

As a side note, I think this post does belong in these threads personally.

Best regards,

Daryl,
Thanks for the chronology; my comment was based upon a MPDS User's Guide one of my clients has from '95, branded by Strategic.

Regards,

Daryl,

I agree with the comments on the criticality and prioritization approach. Really I see it that everything is relative there are no absolutes and therefore some equipment is more important relative to other equipment on site. The goal in my mind was to try and categorize equipment to separate the equipment on site. If everything ends up in the same cell then one is no further ahead.

I assume from your perspective there is no shortcut to take. One must perform the whole FMECA or RCM type process on all their equipment?

My opinion is that it may be necessary for some companies to start by performing a FMECA or RCM process on the more (note that I have used the word more as a relative measure) critical equipment as the resources may not yet be available to perform the analysis on all the equipment on site. I do know that initially some risks will exist with this approach, but I am not suggesting one ends here. I think the program should stay alive and more equipment is analyzed each year, but again this will be the responsibility of each company.

To all,

If a company only has a choice of funding a partial analysis is it better to perform none? By historical cost (probability)? And/or using some “clunky” filter that tries to bring in consequence? Or perform RCM in its entirety on one section of a plant?

I don’t know if there is a definitive answer as most answers will hinge on people’s perspectives, opinions, and past experiences.

A company’s management of change and sustaining processes will also have implications on the value gained.

Jaz,

I am not too sure if I eplained myself correctly. Personally I do think RCM is for everything. Practically thats never going to work, so yes there is a need for some for of targetting excercise.

I dont see this as a shortcut, just a pragmatic approach. What I am suggesting is that instead of ranking equipment based on their criticality, relative absolute importance, you rank them by their importance to todays corporate objectives.

Have a look at these two links (I am not involved with this software in any way apart from being a user once, you can do similar in a spreadsheet)

http://www.expertchoice.com/customerservice/ahp.htm

Or...

http://en.wikipedia.org/wiki/Analytic_Hierarchy_Process

The basic outline of AHP and other prioritization approaches is:

• List the corporate objectives of the organization
  
• Get the decision makers to rank these using a pairwise comparison technique
  
• Analyse and review their responses, taking into account thei consistency of their answers. (E.g. a>b b<c c<a cant be correct logically)
  
• List the benefits of the projects you are doing against each of the coroporate objectives
  
• Use this to create the original priority of analysis projects

This is a pretty brief explanation and the method is far more comple i you want to learn more about it. Not my method, I think it has had some large scale applications in the nuclear industry where it has been supported by court rulings.

In terms of how much of your asset base should you do... it has been my experience that around 20% of the asset base delivers around 60%- 80 % of the benefits. Standard pareto stuff. And then the next 30% will deliver around about the next 10-15%.

Cheers,

Daryl,

Thanks for the response.

I would like to point out that my "clunky" filter is based upon the corporate objectives; perhaps this is not very clear. I did not share the entire key criterion of where I work, but for some HSE priorities can be broken out to which are most important and which situations one would want to avoid. For myself there are still other scoring criteria I did not share. These criteria and the boundaries are from the decision makers to meet the corporate objectives. However my filter would fail to provide any Analytical Hierarchical Process.

I have seen the AHP method before in marketing. For example automobile manufacturers will use the method to survey a sample population to discover what are people's priorities: Power, Fuel Efficiency, Cargo Volume, Occupancy, etc. These criteria are considered relative to each other and after “scoring” this will hopefully provide the manufacturer with the most important qualities they must try to meet and speak to. And these qualities will change based on external environmental factors - such as the high price of oil or demographic shifts to smaller family sizes. As the wikipedia article suggests a certain “orthogonality” to the criterion is important.

I had not thought of applying this methodology to better select equipment for analysis – it is intriguing, but you mentioned something in a previous thread about “not my method…” could you elaborate?

Dear Jaz,

By "not my system" I mean I didnt come up with it, I just apply it.

I have used this to determine the sequence for RCM analyses, determine the priority fo capital spending portfolios to half a billion pounds, and use it regularly as a means of rating strategic initiatives in asset transformation programs.

Once I really "got" what the system does and how it does it I have found no end of opportunities to apply it.

It also has the added benefit of involving the senior management of the plant, organization or institute in the day-to-day reliability activities.

Best regards,

Daryl,
You seem to have serious experience in this analytical hierarchy process?
Do you employ it in a consultancy basis for other companies?

Kencj,
The AHP method looks great. Something else I would like to throw in the mix is simply, timing is everything. Choosing which assets to perform RCM on, in the beginning, should have an element of what matters most to the stakeholders at the moment. The first two or three assets you perform RCM on don't necessarily have to be the ones that rank the highest in a quantitative priority or criticality evaluation. Successful implementation of RCM involves a culture change, and I would recommend the following:

• Determine who in management needs to be convinced that RCM is the way to go, and find out what equipment is important to them.
  
• Identify assets whose failure would impact safety AND operations.
  
• Get buy-in up front from all affected groups by communicating what they can reasonably expect to gain from their investment of resources in RCM (basically answer the WIIFM for key stakeholders)
  
• Utilize the expertise of an experienced RCM consultant at least on the first project.

My personal experience has been to talk to lots of people about RCM and get a "feel" for where it would "hit home" for them. If you want RCM to have the biggest impact, it needs to become the way people think. Early success on highly visible equipment can go a long way in this regard. Then you can go about ranking assets and analyzing them in a systematic manner.

Shelley

kencj

I do apply AHP as part of the management consultancy that I offer. I also apply fuzzy logic style criticality analysis.

It really depends on the application. I have had this conversation on this forum previously and was derided for seeing a difference... but as time goes on it is clear to me that there not only is a difference.. but that it is a big difference that counts a lot!

Some of the applications that I have applied AHP to include:

• Capital planning
  
• OPEX planning
  
• Sequencing RCM analyses
  
• Determining where to spend the improvement budget

Again, once I "got it" with respect to how this works and how this can analyses the decisions that need to be taken, there is suddenly no end to the applications for the method.

Best regards,

As a side note, AHP and Fuzzy logic are directly compared and analysed in my new book which should be published within the next 6 months hopefully.