Page 406 - Improving Machinery Reliability

370 Improving Machinery Reliability

3. What could cause each of a multitude of functional failures?
-Since our systems are likely to include instrument, electrical, hydraulic and
computer-electronic functions, do we really have the in-house talent and time
to collect answers?
4. What exactly happens when each of these failures occurs? What are the short-
and long-term consequences? What is the safety and environmental impact,
downtime impact, community perception impact?
-Can we afford to let our “lean” organization divert its attention away from
the urgent day-to-day business? Will they find answers to some of the obvi-
ously subjective questions raised here? How much time can we afford to
spend on getting these answers? Would a consultant know, or would he have
to sit down in lengthy sessions with our own personnel to get at the answers?
5. What can be done to prevent each potential failure?
-Are we prepared to arrange for the considerable interaction between project,
operations, maintenance, technical, and managerial staff that will be neces-
sary to deal with this question? The decisions, actions, and omissions of each
of these different job functions will influence equipment reliability, failure
risk, and plant profitability. Who chairs these meetings, and how much time
are we prepared to allocate to the task?
6. What is the exact cost justification to implement a given failure prevention
measure?
-Do we have the talent to perform calculations with reasonable accuracy? Is
someone already routinely engaged in this work, or will it be necessary to
train a person?
7. What measurements can be made to predict and/or track failure development?
How often should these measurements be made and who should make them?
Who should interpret them?
-Do we have personnel who have knowledge as to how often certain vibration
measurements should optimally be performed on various machines? Do we
know how, where, and when these measurements are typically being per-
formed on equivalent equipment at “best-of-class” facilities around the world?

Prerequisites to RCM
By now the reader will perhaps appreciate why an estimated 60% of U.S. indus-
try’s attempted RCM implementations are being abandoned after a year or two, and
why the affected plants revert back to their original or some other maintenance
strategies. RCM is anything but a quick fix. If the basic problem is lack of training,
then the implementation of proper training is an indispensable prerequisite to the
pursuit of RCM. If, as in a typical refinery, most failures are related to operational or
product-related upsets, it stands to reason that they will occur at essentially random
intervals. Earlier in our discussion, we had made the statement that only equipment
that exhibits wear-out failures will optimally respond to RCM, and that random fail-
ures are best detected by predictive, experience- and instrument-based data gathering
and analysis techniques. It will thus make better economic sense to emphasize pre-
dictive and component upgrade approaches. These will have to be pursued in con-
junction with cost-justified state-of-art work processes, sound and effective zero-