fail, our current maintenance planning process starts directly with the equipment, and its sole purpose is to specify actions required to "keep it running."

But recognize from the outset that Reliability-Centered Maintenance is not just another cleverly packaged way to do the same old thing again.  Rather, it is very different in some fundamental aspects from what today is the norm among maintenance practitioners, and requires that some very basic changes in our mindset must occur.  As you will see in a moment, however, the basic RCM concept is really quite simple, and might be viewed as organized common sense.

So just what is RCM?  There are four features that define and characterize RCM, and set it apart from any other maintenance planning process in use today.  We will use a hypothetical scenario to develop and understand these four features.

Feature 1

Picture a typical business conference room which, we will hypothesize, represents the location of a system in our process plant.  As we stand outside the walls (i.e., boundary) of the room, we observe that a 24 inch diameter pipe is moving water at ambient pressure and temperature into the room (i.e., system).  At the other end, we find a 24 inch diameter pipe exiting the room, but now the water has been elevated in pressure and temperature to some levels that are required elsewhere in our process plant.

Notice that at this point, we (theoretically) have no idea what is inside the room.  But whatever this may be, it has made the room capable of elevating water pressure and temperature.  We call this capability the function of the room (i.e., system), and we are able to accurately define this function without any knowledge of the room contents (i.e., equipment).  In order for our plant to produce its end product, we must assure that this system continues to perform its job.  That is, we must "preserve system function"-and this is the first and most important feature of RCM.

At first glance, this is a difficult concept to accept because it is contrary to our ingrained mindset that PM is performed to preserve equipment operation.  By first addressing system function, we are saying that we want to know what the expected output is supposed to be, and that preserving that output (function) is our primary task at hand. 

This first feature enables us to systematically decide in later stages of the process just what equipments relate to what functions, and not to assume a priori that "every item of equipment is equally important," a tendency that seems to pervade the current PM planning approach.

Feature 2

Since the primary objective is to preserve system function, then loss of function or functional failure is the next item of consideration.  Functional failures come in many sizes and shapes, and are not always a simple "we have it or we don't" situation.  We must always carefully examine the many in-between states that could exist, because certain of these states may ultimately be very important.

The paramount question, then, would be to ascertain just what has happened inside the room to produce the functional failure.  To answer this question, we now open the door and step into the room (system).  There before us are all of the components (equipment) that are working together in some harmonious manner to produce the function that was observed when we were standing outside the walls (boundary) of the room.  Our job now is to meticulously examine each component in order to delineate just how it might fail such that the functional failure(s) could occur.  Thus, the key point in Feature 2 is that we make the transition to the hardware components by "identifying specific failure modes that could potentially produce the unwanted functional failures." By way of illustration, a flow control valve (component) that is jammed shut (failure mode) could produce the functional failure "fails to initiate system startup."

Feature 3

In the RCM process, where our primary objective is to preserve system function, we have the opportunity to decide, in a very systematic way, just what order or priority we wish to assign in allocating budgets and resources.  In other words, "all functions are not created equal," and therefore all functional failures and their related components and failure modes are not created equal.  Thus we want to "prioritize the importance of the failure modes." This is done by passing each failure mode through a simple, three-tier decision tree which will place each failure mode in one of four categories that can then be used to develop a priority assignment rationale.  (This will be discussed in detail in a subsequent article.)

Feature 4

Notice that, up to this point, we have not yet dealt directly with the issue of any preventive maintenance actions.  What we have been doing is formulating a very systematic roadmap that tells us the where (component), what (failure mode), and priority with which we should now proceed in order to establish specific PM tasks-all of this being driven by the fundamental premise to "preserve function." We thus address each failure mode, in its prioritized order, to identify candidate PM actions that could be considered.

And here, RCM again has one last unique feature that must be satisfied.  Each potential PM task must be judged as being "applicable and effective." Applicable means that if the task is performed, irrespective of cost, it will in fact accomplish one of the three reasons for doing PM (i.e., prevent or mitigate failure, detect onset of a failure, or discover a hidden failure).  Effective means that we are willing to spend the resources to do it.

Generally, if more than one candidate task is judged to be applicable, we would opt to select the least expensive (i.e., most effective) task.  In describing a run-to-failure task category, there are three reasons for such a selection.  To be more precise, failure of a task to pass either the applicability or effectiveness test results in two of the run-to-failure decisions.  The third would be associated with a low-priority ranking and a decision not to spend any PM resources on such insignificant failure modes.

In summary, the RCM methodology is completely described in four unique features:

1. Preserve functions.
2. Identify failure modes that can defeat the functions.
3. Prioritize function need (via failure modes).
4. Select applicable and effective PM tasks for the high priority failure modes.

The above four features totally describe the RCM concept —nothing more and nothing less.  For any maintenance analysis process to be labeled as RCM, it must contain all four features.  The authors have occasionally encountered maintenance programs that are purported to be RCM programs but lack one or more of the four features.  And usually these programs are also less than satisfactory, and tend to give RCM an unfair reputation.  So we caution you to avoid the shortcuts if you truly wish to have an RCM-based PM program.

Printed with permission from Butterworth-Heinemann, a division of Elsevier, from RCM - Gateway to World Class Maintenance, by Anthony M. Smith, AMS Associates Inc. in California, and Glenn R. Hinchcliffe, Asset Performance Technologies Inc., Corrales, NM.  Copyright 2004.  For more information about this title and similar titles, please visit www.books.elsevier.com.