seven steps that have been developed from experience as a most convenient way to systematically delineate the required information:

Step 1:  System selection and information collection.

Step 2:  System boundary definition.

Step 3:  System description and functional block diagram.

Step 4:  System functions and functional failures-Preserve functions.

Step 5:  Failure mode and effects analysis (FMEA)-Identify failure modes that can defeat the functions.

Step 6:  Logic (decision) tree analysis (LTA)-Prioritize function need via the failure modes.

Step 7:  Task selection-Select only applicable and effective PM tasks.

Satisfactory completion of these seven steps will provide a baseline definition of the preferred PM tasks on each system.  Notice that Steps 4, 5, 6, and 7 correspond to the four basic features that characterize RCM.  Two additional steps are required in order to complete a successful RCM Program:

Step 8:  Task packaging-which will carry the recommended RCM tasks to the floor.

Step 9:  Living RCM program-comprising the actions necessary to sustain over time the beneficial results of Steps 1-8.

Key points to remember

When an analyst embarks upon the process described here, it is helpful to keep a few key points in mind relative to the application of the RCM systems analysis process.

1.  Traditional methods for determining PM tasks focus the entire task selection process on what can be done to the equipment.  As a rule, why it should be done is never clearly addressed.  RCM is a major departure from this traditional practice.  Its basic premise is "preserve function"-not "preserve equipment." This approach forces the analyst to systematically understand the system functions that must be preserved and then to think carefully about how functions are lost-in functional failure terms, not equipment failure terms.  The purpose of this approach is to develop a credible rationale to perform an appropriate PM task rather than just arbitrarily deciding to do something because "it sounds right." (The "preserve function" approach is initially developed in Steps 3 and 4.)

2.  However, this is not to imply that traditional experience and sound engineering judgment about equipment malfunctions is unimportant to the RCM process.  On the contrary, the use of operations and maintenance personnel experience, as well as historical data from plant-specific data files is an invaluable input to assuring that all important failure modes are captured and considered in the FMEA (Step 5).

3.  The direct involvement of plant operations and maintenance personnel in the RCM systems analysis process is extremely important from another point of view also-namely, as a "buy-in" to the process.  Experience with many RCM programs has shown that success is rarely achieved if the buy-in factor has been neglected.

4.  Recall there are four categories from which to choose a candidate PM task: (1) time-directed (TD), (2) condition-directed (CD), (3) failure-finding (FF), and (4) run-to-failure (RTF).  As a rule, there is virtually no difficulty with people accepting the definition and use of TD and CD tasks in a PM program.  Use of the FF task as a formal inclusion in the PM program is new to most people, but is generally accepted as a valid PM task in a short period of time.  But the notion of a deliberate decision to run-to-failure is totally foreign to the more traditional elements of preventive maintenance.  Thus, some care and sensitivity to the use of RTF tasks may entail some special education efforts to ensure that the operations and maintenance personnel understand why RTF tasks are, in fact, the best selection.  The specific reasons behind RTF are developed in Steps 5, 6, and 7.

5.  It is not uncommon for people receiving their first exposure to Classical RCM to comment that "there sure is a lot of paperwork involved here." And there is some truth in the comment.  Thus, it becomes important to help people to understand why the paperwork is there, and how it benefits them in the long run.  These points should include the following:

"RCM wants to ensure that you can answer, both today and in the future, the "why" behind every task that will use your limited resources (i.e., preserve the most important functions).

"RCM wants to ensure that your task selections derive from a comprehensive knowledge of equipment failure modes because it is at that level of detail where failure prevention, detection, or discovery must occur.  If your task selection process does not do this, then there is no assurance that the task really does anything particularly useful.

"RCM wants to ensure that the most effective (least costly) task is chosen for implementation.  Historically, this has not been done and, consequently, most PM programs fall far short of realizing the best return for the resources spent.

In order to realize these benefits, it does take some effort and documentation.  But, once a system has been through the RCM process, it produces a baseline definition of the PM program for that system which needs only periodic update to account for new information and system changes.  Further, as the RCM process has evolved and matured, much of the mechanics of the analysis has been computerized, and this has introduced efficiencies as well as eliminated the need for hard copy reports where such was desired.

6.  It should be noted that the RCM methodology focuses only on what task should be done and why (i.e., task definition).  All tasks must likewise establish when the task should be done (i.e., task frequency or periodicity), but these intervals are derived from separate analyses that must consider and utilize combinations of company and industry experience to establish initial task frequencies.  More sophisticated statistical tools may be employed when data is available to pursue this avenue; also, controlled measurement techniques known as Age Exploration can be used.

7.  If possible, the systems analysis process should involve a team of two or three analysts and one of them should be from operations.  This will encourage not only cross-talk about what information should be included in each of the seven steps, but also a healthy level of challenge, questioning, and probing in that regard. 

8.  Last, it is important that proper consideration be given to scheduling the activities of the RCM team in order to minimize any impact on the normal responsibilities of the team members.

Experience has shown that the analyst will be placed in a more proper state of mind to proceed with an RCM systems analysis if some initial appreciation of the above points has been acquired and accepted.

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.