Page 33 - Intro Predictive Maintenance

24 An Introduction to Predictive Maintenance

2.1 ASSESSING THE NEED FOR CONDITION MONITORING
Any maintenance engineer’s assessment of plant condition is influenced by a variety
of practical observations and analyses of machine performance data, such as the
following:

• Frequency of breakdowns
• Randomness of breakdowns
• Need for repetitive repairs
• Number of defective products produced
• Potential dangers linked to poor performance
• Any excessive fuel consumption during operation
• Any reduced throughput during operation

These, and many more pointers, may suggest that a particular item of plant requires
either careful monitoring, routine planned preventive maintenance, better emergency
repair procedures, or some combination of all these approaches to ensure a reason-
able level of operational availability. The engineering symptoms can, however, rarely
be quantified accurately in terms of financial loss. Very few companies can put an
accurate figure on the cost of downtime per hour. Many have no reliable records of
their aggregate downtime at all, even if they could put a value per hour on it.

Thus, although a maintenance engineer may decide that a particular machine with a
history of random bearing failures requires CM, if problems are to be anticipated, and
the plant should be taken out of use before a catastrophic in-service failure occurs,
how can he or she justify the expenditure of, say, $10,000 on the appropriate moni-
toring equipment, when plant and production records may be too vague to show what
time and expense could be saved, and what this savings represents in terms of profit
and loss to the company? This dilemma can be a daily occurrence for engineering and
maintenance staffs in large and small companies throughout the country.

As if the practical problems of quantifying both the potential losses and gains were
not difficult enough, the status of maintenance engineering in many organizations is
such that any financial justification, however accurate, can be meaningless. The main-
tenance department in most companies is usually classified as a cost overhead. This
means that a fixed sum is allocated to maintenance each year as a budget, which covers
the cost of staff wages, spare parts, consumable items, and so on. The maintenance
department is then judged for performance, financially or on its ability to work within
its budget. Overspending is classified as “bad,” and may result in restricting the depart-
ment’s resources even further in future years, whereas underspending is classified as
“good,” in that it contributes directly to company profits, even if equipment mainte-
nance is neglected and manufacturing quality or throughput suffers as a result.

Let us suppose that a forward-looking engineer succeeds in persuading his or her
financial director—who knows nothing about CM and would rather invest the money
anyway—to part with the capital needed to buy the necessary CM equipment. What