Preventing Failures of RCM - Pt 1

After a long and winding journey RCM is finally findings its rightful place as a cornerstone of modern asset management. The benefits are well documented now and cover all aspects where physical aspects have an impact on corporate performance.

Yet still many programs end with a whimper.

Lack of management support, poor asset selection, lack of momentum and taking technical shortcuts are undoubtedly killers of any RCM program. The overuse and misuse of criticality, streamlining the method instead of the implementation process and poor program management accounts for a lot of these issues.

Yet all are joined in one classic error; the failure to adequately train RCM Analysts.

When reviewing RCM analyses I often come across annoyingly similar mistakes, all of which have potentially harmful impacts, and all of which are avoidable if the RCM Analysts are properly prepared in the first place.

These errors tend to fall into three categories, and aside from the impacts below they are all motivation and momentum killers.

Failure modes at the wrong level of causality. Leads to blanket strategies not connected to the failure mechanisms, over use of the run to failure options, and excessive spares options required.

The worst effect of this of course is the fact that not all of the reasonably likely failure modes have been uncovered. Yet everyone familiar with RCM expects that they are. False sense of security, unmet expectations, classic failure of the implementation.

Combed of ambiguous failure modes - "Control Failure" is a classic example of this, so too is "bearing fails due to wear or contamination". Again there is almost no way to get the right sort of failure management strategies in these situations. This is not uncommon where someone is trying to justify what already exists, instead of determining the real maintenance requirements.

Developing strategies without using the decision diagram(s) - This is a shocker and almost always happens to first time RCM Analysts. The trap is to fall into inserting the strategies that already existed instead of developing strategies to manage the failures found. This has the effect of a zero benefit analysis, or worse - one that is incorrect. (And potentially dangerous)

It is easy to come undone here. Time based failures really seem like they should be managed using a predictive task of some sort... but how is this done? Lots of questions like this exist between completing an RCM Analysts course and becoming good at it.

Misapplication of the Detective maintenance formulas - This is far too big to discuss as part of this stream. But suffice to say that it is one of the real potentially dangerous areas of the analysis.

These are all errors related to technical soundness of the analysis, and there are of course, lots more like:

• Defining design capacity instead of user requirements
• Defining functions that are already covered by the primary function
• Basing detective maintenance frequencies on evident (safe detected) instrument failures... and so on...

In the next posting in this series we will go into some of the program management failures. Those failures that almost guarantee the work will get little support, little momentum, and very little chance of accomplishing what the organization has set out to accomplish. 

The way to combat these failures is to have adequately trained RCM Analysts. Analysts who have received both classroom and on the job training, as well as being coached through the program management issues.