 The
role of PdM in asset care
A technology that tells you what you need to do
A complete, accurate
knowledge of the critical production or manufacturing systems that comprise
the capital assets of a plant must be the first requirement of any asset
care program. Predictive maintenance is the best tool for this critical
job. Used properly, predictive maintenance technologies provide on-going
analyses of actual operating condition, the motivation for preventive or
corrective maintenance, and identification of the remaining useful life
of these plant systems. This primary role of predictive maintenance has
several aspects in asset care.
Benchmarking plant systems
There is a misconception that predictive maintenance is only useful
for long-term trending of equipment condition. Nothing could be further
from the truth. Predictive maintenance technology is an ideal benchmarking
tool. While the long-term condition history of plant systems is the cornerstone
of asset management programs, these technologies are not limited to this
approach.
In many cases, the production capacity gain was the
direct result of elimination of unnecessary planned downtime.
As in every use of predictive maintenance, the design specifications of
each machine-train and production system is the key to its proper use.
From this data, one can calculate the operating dynamics, including vibration
profiles that the system should produce in normal operation. In addition,
this approach should evaluate each system to define design weaknesses,
the most likely failure modes, and its normal range of acceptable operation.
Operating procedures can be checked to make sure they coincide with the
designed operating parameters. In other words, we analyzed the machine-train
or system so that we get a full understanding of how it was designed to
operate.
With this data, one can measure the actual operating dynamics of each
machine-train and system. Engineers can use time- and frequency-domain
vibration, thermographic scans, tribology, and a variety of other predictive
tools to measure the specific parameters that define the actual dynamics
and condition of production system during normal production. It is imperative
that the data include the full range of normal operating conditions, including
ramp-up and deceleration on variable speeds systems.
The data, recorded under the full range of normal operating conditions,
quantifies the system's operating condition, system reliability, and
projected remaining life. The data identifies deviations from optimum
operating conditions as well as isolates any incipient problems that may
be present within the system.
The benchmark evaluation is inclusive. It identifies and isolates problems
that result from design limitations, operating practices, maintenance practices,
and product variations. The information derived from a comprehensive benchmark
analysis provides the starting point for asset care management. In addition,
it provides the information for cost-effective corrective actions that
can be implemented as part of the asset care program to ensure optimum,
long-term performance from each production system.
As in every use of predictive maintenance, the design
specifications of each machine-train and production system is the key to
its proper use.
Maintenance scheduling
Data derived from predictive maintenance programs provides the most
accurate means of scheduling corrective and preventive maintenance tasks.
When used properly, predictive maintenance defines the specific interval
and type of maintenance tasks required to sustain optimum operating condition
and to extend the useful life of critical plant systems.
Too many plants use the data derived from their predictive maintenance
program as an adjunct to their normal time- or history-driven scheduling
practices. As a result, they continue to make unnecessary repairs that
not only increase maintenance costs but also decrease the useful life of
plant systems. Predictive maintenance data should be the dominant, if not
the only, driver for maintenance scheduling. Comprehensive use of vibration,
tribology, thermography, and other techniques provides a more accurate
means of scheduling tasks needed to maintain optimum system health.
Maximum equipment utilization
These same predictive maintenance techniques can be used to achieve
maximum utilization of plant resources. Regular evaluation of equipment
condition and prompt attention to deviations detected by the predictive
analysts eliminate both scheduled and unscheduled downtime. Coupled with
an improvement in overall reliability, predictive maintenance radically
increases the useful production time of these plant systems.
Over the past few years, we have used predictive maintenance to gain
substantial increases in first-time-through capacity. In many cases, the
production capacity gain was the direct result of elimination of unnecessary
planned downtime.
In addition, the knowledge gained from predictive maintenance helps
production planners determine the actual capacity limitations of critical
plant systems. With this knowledge, the planners are in a much better position
to schedule production runs to achieve the maximum utilization of plant
systems.
Factory and site acceptance testing, using predictive
maintenance technologies should be an integral part of procurement.
New equipment specifications
Predictive maintenance can be used to develop valid specifications
for new production systems and equipment. Historical data developed from
existing systems can be used to develop specifications that ensure reliability,
maintainability, and life-cycle cost on new systems. Every viable predictive
maintenance programs should include a comprehensive machine history that
defines the failure modes,
frequency of failures, and other vital data that can be used to develop
new equipment specifications.
In addition, the procurement process on every new equipment item should
include a comprehensive set of tests that ensure proper selection and installation
of the new equipment and systems. Factory and site acceptance testing,
using predictive maintenance technologies, should be an integral part of
procurement.
Over the past few years we conducted numerous site acceptance tests
for our clients. In more than 50 percent of these tests, we found serious
problems that would have reduced the reliability, maintainability, and
life cycle costs of these new systems substantially. The sources of these
problems are in two major classifications: deviation from material specifications
and improper installation. In each case, these problems were corrected
before final acceptance of the systems.
Upgrade system performance
Historical data, developed as part of a predictive maintenance program,
should be the basis for upgrades to critical process equipment. Too many
plants implement upgrades or redesign of process systems without adequate
consideration of their impact on reliability. In many cases, these upgrades
result in a substantial reduction in availability and an increase in operating
costs. More of our clients are using these data for the design and implementation
of system upgrades.
In addition to the historical data, many plants are using traditional
predictive maintenance techniques such as vibration analysis to determine
the effect of process modification before the upgrade design process begins.
This approach results in substantial improvements in capacity, product
quality, and overall life-cycle costs.
Without the complete involvement and use of a comprehensive predictive
maintenance program, asset care management has little chance of success.
Since its prime function is to maximize the benefits of critical plant
assets, every task and activity of the asset care program must be driven
by the actual condition of these assets. Predictive maintenance provides
a cost-effective means of obtaining this essential knowledge.
Copyright June 1998 Plant Services on the WEB
|
