Common Errors Implementing RCM

When carrying out an RCM analysis I often get a look at past efforts to perform some form of strategy review. Yet asset performance remains inconsistent. (At best!)

Below is a list of common errors I find in these sorts of analyses. This is not a trivial issue. Failed strategy programs produce failed maintenance, and failed maintenance produces assets that are unreliable, with high likelihood of failures leading to operational, safety or environmental integrity consequences.

I hope these are useful for you. I would also want to counter some of the uninformed hype out there that any RCM is good for you.

If it is poorly done then it can quite easily put you further behind where you are now!

#1 Error: Failure modes not relevant to the asset or the operating conditions of the plant. This is exceedingly common.

Asset configuration, type, location and industry sector are all issues here. A gold mining conveyor operates in a far different environment than a coal mining conveyor. And both are different to an underground conveyor.

For example, failure modes for a duty pump listed against a stand by pump; or no failure modes listed to take in the fact that this conveyor (say) operates on the sea front.

Cause: Indiscriminate use of templates without checking against the operating context.

Danger #1: Un-managed failure modes.

One of the key reasons why asset analyses and asset performance seem to be unrelated in many cases. This can be incredibly dangerous because human error style failure modes are often a large percentage of those that are left out.

Danger #2: Over maintenance caused by irrelevant failure modes.

This can be very dangerous. First, it often puts people in sometimes hazardous situations without any requirement. And second, because they get ignored!!

The people carrying these work instructions out are not stupid. They know when they don't make sense, and they avoid wasting their time. This means that if there are any important strategies hiding there, that really need to be done, they will be thrown out with the garbage.

#2 Error: Failure mods listed at the wrong level of causality. (Includes use of generic terms like breaks/fails)

This is a killer, literally. For example, listing a failure mode as bearing fails, when the real failure modes come from errors during installation.

What would happen here? The team would recommend VA, and it WOULD capture the early signs of failure. protecting us from unplanned failure consequences. Yet the bearing would have had a short life, and this would continue as there were no plans or strategies in place to take care of that failure mode.

Cause: Skill levels and asset knowledge of the facilitator / analyst and of the people carrying out the review.

Danger: Ineffective maintenance is the real danger. Maintenance strategies that will not address the failure modes that the machine is likely to experience.

The result is continued poor performance and rising maintenance costs through early life failure.

#3 Error: Turing RCM into an exercise to populate a database instead of correcting asset performance.

Cause: Software focused implementation. Inexperienced analysts, data instead of performance focus.

Danger: All of the above, PLUS. This is incredibly dangerous from an asset performance and from a safety point of view. Yet it is exceedingly common.

When the analysis effort is divorced from the asset performance outcome nothing at all good can come from it!

#4 Error: Ineffective and inapplicable maintenance.

Cause: Lack of a decision diagram or incorrect use of a decision diagram.

Danger: Applying the wrong maintenance leads to many potential hazards. Among them are wasted efforts, irrelevant regimes, higher likelihood of human error and a range of others.

In short this is the surest way to guarantee failure of your maintenance function.

#5 Error: Omitted failure modes 

Cause: Improper listing of functions and / or functional failures

Danger: Omitted failure modes means that there will be failures, and their consequences, that will occur because your team was not able to even identify them let alone manage them. 

Summary

Beware of anybody bearing news of faster methods of implementing RCM. There are rapid ways to do it, and they are very effective. But make sure that they cover the basics.

They must list the functions and the functional failures in order to ensure that ALL reasonably likely failure modes are identified.

They must list ALL reasonably likely failure modes at the right level of causality. (The level where you can do something proactive about it), and

They must select the failure management strategies in line with the effectiveness and applicability criteria contained within the RCM decision algorithms.

Good luck!

If you enjoyed this post you may like to subscribe to get each blog post as they are written in your inbox here.