Reliability – Centred Maintenance
The traditional approach to scheduled maintenance starts
by defining a list of maintenance tasks, consisting of both
inspection and repair or replacement activities; these are
then built into schedules, the frequency with which they
are applied being specified in terms of either elapsed time
or time in use.
RCM in contrast, starts by considering what would happen
if you didn’t undertake such tasks, i.e. what could cause
a failure of the equipment, and what would be the consequences
of each of these possible failures. This means that you
can get better value for money from your scheduled maintenance
activities, since with RCM these are carefully targeted:
if the initially failure modes and effects analysis indicates
that the consequences of failure are minimal (as is often
the case), then no preventive action is justified; in other
words, RCM accepts that there are occasions when the old
‘if it ain’t broke, don’t fix it’ policy can be the most
cost-effective approach. You could well find that this results
in getting on for half your previous scheduled maintenance
tasks being eliminated: this saves not just the labour cost
involved, but also the cost of all the spare parts that
would have been fitted ‘just-in-case’, as part of the scheduled
maintenance routine.
To illustrate the point, consider what scheduled maintenance
would be required on a simple water pump. The old way would
be to get a drawing of the pump and make a list of parts
such as bearings and seals which need to be checked regularly
and replaced periodically; if there are several identical
pumps used in the factory, that makes the task easier, because
the same list of maintenance tasks is sued for each pump.
The RCM way, in contrast, is to say, ‘Hang on minute, where’s
the pump going to be used?’ If the answer is ‘In the nuclear
reactor cooling system’, then the effects of failure could
be so serious that some form of protective maintenance would
be considered essential. But suppose the answer was ‘To
keep the water circulating in the ornamental pond outside
the main reception area’, and then it would probably be
cheaper to leave it alone and accept the risk that it might
fail. In other words, an RCM schedule takes full account
of the operating context of the equipment and of the consequences
of failure.
That raises another question: what constitutes ‘failure’?
Suppose your car works perfectly well in all respects except
that every time you exceed 80 k.p.h it has a tendency to
jump out of gear: but the car will still get you to your
destination. Is this just a sign of a potential gearbox
failure that could eventually lead to an actual failure
at some future date so that you don’t reach your destination;
or has it already failed because you are unable to use the
car to its maximum potential? This sort of problem happens
all too often in a factory, when production equipment is
routinely operated at less than its rated output speed,
because the operators know from experience that problems
will occur if they try to run the equipment any faster.
In world-class terms, a ‘waste’ occurs if you are unable
to use equipment to its full potential and, as you know
by now, you should endeavour to eliminate or, at least,
minimise any form of waste. Accordingly the RCM definition
of failure is:
With these underlying principles explained,
I can now summarise the key stages in the RCM procedure:
-
Start by defining, for each piece of equipment,
how it might fail and what affects each failure mode might
have.
-
Use the RCM Decision Diagram to assess
how critical these effects are. Do they affect personal
safety, so that the risk of failure must be avoided altogether,
if necessary by redesign? If personal safety isn’t affected,
what would be the economic consequences of failure? Does
the cost penalty if failure occurs justify the cost of
preventive measures?
-
For each safety-critical failure mode,
decide what action must be taken to overcome the risk
of such a failure occurring.
-
For each failure mode with significant
economic consequences: decide whether the failure pattern
is age-related, so that a ‘scheduled discard’ task can
be planned: if not, what ‘potential failure’ signs could
be used to indicate when corrective action is needed,
or what condition monitoring techniques or on-condition
maintenance tasks might be appropriate.
-
For the remaining failure modes, i.e.
those defined as ‘tolerable’, decide what action can be
taken to minimise the risk of them occurring (for example,
by getting operators involved in TPM in order to improve
equipment reliability).
To sum up: RCM enables you to target your maintenance resources
precisely in order to minimise waste, whether this be the
waste of equipment breaking down, or the waste of equipment
working below par, or the waste implicit in carrying out
‘preventive’ maintenance work that has no real effect. Whether
the amount of work involved in a full RCM application is
worthwhile will depend on your particular circumstances.
However, it will nearly always be worth applying the basic
concepts to at least one or two key machines; sufficient
information to enable you to do this, and to decide what
degree of detail is appropriate.
