In RCM, if a component fails and nothing happens, is it automatically classified as a run to failure component?

Terry O

Yes; but that is a non-preferred method to discover RTF components.

This isn't like, if a tree falls in the forest and no one is around to see or hear it, will it make a sound?

I not exactly sure of the question. But then again, there's a number of things I'm not sure of.

Did the operation fail or shut-down? Was the component an element that caused operation to cease? If operation continued and planning and scheduling could be accomplished, then I would not classify it as a run-to-failure case.

But, I classify run-to-failure as not having a program at all.

Terry meant there was no operational, safety, or environmental failure consequences when the component failed.

IMO there are three classifications of RTF components:
1. RCM analysis identified the component as a critical failure, but LTA pointed to RTF as the most cost-effective PM.
2. Equipment scoping ruled out RCM analysis for the component.
3. Terry’s example, where no effort has been made to identify a significant equipment list / conduct RCM.

Regards,

Terry,

There is never a time when "nothing happens". It may have no safety, environmental or direct operational consequences but it will have non-operational consequences. (Cost of labour and repairs to get it up and running again)

Choosing an appropriate task will therefore depend on the applicability criteria firstly (is it physically possible to do the task?) and the effectiveness criteria secondly.

If a task is technically feasible,(and we wil say that it is) then the question becomes one of cost effectiveness.

Is it more cost effective, over a reasonable length of time, to do the maintenance or to run-to-failure?

This will determine whether you do nothing or not.

Even when the consequences are only repair costs, it may be more cost effective to perform maintenance than to RTF. (More often than not, however, it is not)

IMHO ruling out maintenance prior to taking a failure mode through this step could lead to slightly lower cost effectiveness, particularly if this analysis is to be used as a template for future adaptations. (In which case it will be magnified)

My 2c worth.

quote:

Originally posted by Ozgipsy:
IMHO ruling out maintenance prior to taking a failure mode through this step could lead to slightly lower cost effectiveness, particularly if this analysis is to be used as a template for future adaptations. (In which case it will be magnified)

Daryl,
Are you suggesting all components must go through FMECA without first screening out those that are functionally non-significant?

Armed with a well-thought out exclusion criteria, an analyst can eliminate entire classes of components that would otherwise require (expensive) time to needlessly evaluate. For example; classifying manually operated drain valves, 0.5” or less, as non-significant could eliminate thousands of components from RCM analysis.

Regards,

Larry,

I am not saying that at all. Asset Selection is another issue and not what we were talking about here I didnt think.

However, once selected I suggest that all reasonably likely failure modes are analysed through the RCM process. In the hands of a competent analyst, as you know, this is not a timely task.

That doesnt include a FMECA at all.

I would also suggest that generally it is more time consuming to focus an analysis at the component level, and that an analysis focussed at the system / process level would be a more efficient use of time.

Using a comprehensive asset selection and clasifying process, coupled with a template definition process, can enable companies to determine which asset systems are relatively similar to allow for rapid adaptation of templates with minimal time.

I think this dovetails in with what you are saying.

quote:

Originally posted by Ozgipsy:
I think this dovetails in with what you are saying.

Hi Daryl,
Thanks for the quick response - I didn't know you were on-line right now.

quote:

I think this dovetails in with what you are saying.

I think we are on the same page, but what do you mean by template definition? We’ve been using PM Templates to augment/supplement LTA for some time now. Ours basically recommend a PM strategy based upon component type/class and their duty, environment, and failure criticality. Is this the same as your template definition?

Terry:

It's all about how much it hurts when the particular failure mode occurs that drives the decision to maintain or not to maintain.

I will say that in the majority of cases where there are no safety, environmental or operational consequence it does result in a run to fail decision. Even though there may be a perfectly feasible task, the decision to create a maintenance task or not should be based on the economics of the failure mode / cost of repair as Daryl pointed out.

Assessment of the economic impact for these failure modes is often a step in RCM that is given only cursory attention or is bypassed completely. I believe it is important that people understand the difference between technical feasibility and economic viability if we are to really move maintenance paradigms.

A lot rests on the definition of non-operational consequences. I've heard it argued that every failure has an operational consequence. I can't say that I agree with that point of view 100%.

"If a component fails and nothing happens" I have seen this on a few occasions - the questions I'd ask are: Why was it there in the first place? Do we need it? Can we remove this superfluous item from the system?

Delboy

Larry,

I have two kids and demanding clients... I never sleep!

I don't think we are a million miles away, although there are some particular twists that we use in template classification, application and usage.

I would rather discuss this off line if that okay with you. darylm@strategic-advantages.com

Terrence,

Delboy is spot-on. The only purpose of maintenance is to eliminate or reduce the consequence of failure to an acceptable level. This is a paradigm shift, as in the past(pre-N&H), we believed we had to do something with respect to any failure.
It is perfectly OK to let very low consequence failures to happen, and run-to-failure is a perfectly respectable strategy under the right conditions.

I think there is an important aspect of Terry's question that was overlooked.

I would agree that if a failure occurs to to a component or system(either partial or total) and there are no safety or environmental consequences and the cost of the PM task is greater then the operational cost of accepting the failure then RTF is an exceptable strategy only if the failure of the component or system is evident to the operators of the equipment (under normal operating conditions). If the failure is not evident (ie hidden)and requires an addition failure or plant event to become evident,then the consequences of the component or system being in a failed state need to be evaluated assuming that the event that causes them to operate has occurred (ie plant mode or additional failure). In other words, hidden failure modes need to evaluated assuming a multiple failure scenario.

Is run to failure a planned or unplanned breakdown maintenance?

James is absolutely spot on, I missed the 'evident' aspect.

Josh, RTF is a planned activity.

RTF: Plan to do nothing until fails.

One of the rare occasions when "do nothing" is the right thing to "do".

When we say nothing happens on a run to failure mode of an asset then the consequences of failure must have been found to be negligible. The bigger problem I see is most companies run many of their assets to failure because they do know where on the PF Curve they are operating. PF Curve? what is that?

Ricky Smith, CMRP

PF curve ? Ok I give what is PF curve?

Mike,

I attached a slide that will explain the PF Curve it very simply. If not let me know. Would love to hear comments on it.

PF_Curve1.PPT (785 Kb, 74 downloads) PF Curve Explanation

Ricky, If I understand the curve right,
* Run to Failure strategy is to fix after the F point in the curve, but hopefully before "equipment broken" point? or also after the "equipment broken" point?
* PM strategy is to act timely based before the Failure start point, then shifting the curve forward in time so the green zone extends to the future?
* Predictive Maintenance strategy is to inspect the equipment to predict when the P-F Interval is to schedule a planned corrective before the F point?

What is the formula for the P-F curve?