Chemical Engineering March 2023 - 40

other cases.
Solutions. Before implementing a
predictive-maintenance strategy,
chemical plant owners should carefully
evaluate what can and can't be
predicted. They should only tackle
those tasks that will provide significant
value and avoid trying to make
the impossible possible.
By doing this, they can create a
best-in-class predictive-maintenance
program that will optimize the chemical
plants' equipment and overall
equipment efficiency (OEE).
2. Erroneous equipment selection
and failure to update the system.
Predictive maintenance can take
chemical plant operations to the next
level. However, the plant must have
a sound maintenance-management
system (MMS) in place to be effective.
An MMS is a crucial factor in
achieving an efficient predictive
maintenance program. It enables
the collection and analysis of data
so that CPI plant owners can identify
trends. It also allows for the execution
of planned maintenance tasks
instead of relying on unplanned corrective
maintenance.
However, an MMS is only effective
if it is up to date. Outdated data can
lead to erroneous equipment selections
and prevent the plant from getting
the full benefits of a predictive
maintenance program.
It is also important to note that not
every piece of equipment in a CPI
plant needs to be monitored by predictive
maintenance. The equipment
selected for monitoring should be
based on the plant's specific needs
and the types of failures that are
most likely to occur. A good starting
point is to identify the critical assets
in the plant and focus on those first.
Solutions. To get the most out of
a predictive-maintenance program,
chemical plant owners must ensure
that their MMS is up to date and accurate.
They should also carefully
select the equipment that will be
monitored and focus on their critical
assets first.
3. Employing the wrong personnel
and inadequate training. Employing
the wrong personnel is another
common mistake. This can happen if
the required skill sets do not exist inhouse
or predictive maintenance is
used to cut training costs. The best
predictive maintenance programs
use a combination of highly skilled
40
personnel in their respective fields.
Such a team should be able to work
together to develop and implement
the program.
If there isn't the right personnel
in-house, the chemical plant manager
may be tempted to cut costs
by training existing staff members.
However, this can often lead to inadequate
training. The worst part about
inadequately trained staff is that they
can do more harm than good.
Solutions. To avoid hiring the wrong
and inadequately trained staff,
the
management can do the following:
* Hire personnel with the right skillset
* Ensure that team is adequately
trained
Hiring the right personnel. There are
a few things to keep in mind when
hiring predictive-maintenance personnel.
The first is that they should
have experience in their respective
field. This experience can come from
working with similar machinery or
predictive maintenance software.
Second, they should be able to
work together as a team. Predictive
maintenance is often a complex
process requiring several different
experts' input. Third, they should be
able to communicate effectively -
this means understanding and explaining
complex concepts so that
others can understand.
Fourth, they should be able to
train others. As mentioned earlier,
one of the main tasks of predictivemaintenance
personnel is to train
staff members on using the software
and equipment.
Ensuring adequate training. To ensure
that staff is adequately trained,
management has to do the following:
* Set clear training goals
* Make sure that the training is
interactive
* Use simulations where possible
* Make sure that the training is relevant
to the job
One of the best ways to ensure
that staff is adequately trained is to
set clear training goals. For example,
a goal for staff members could be
the ability to operate the predictivemaintenance
software by the end of
the training session.
4. Not analyzing and integrating
collected data. It isn't efficient to
just set and forget a predictive-maintenance
program - data analysis is
a crucial part of it. After all, the insights
gleaned from data will help
fine-tune the program and prevent
issues down the line.
Analyzing data can be daunting,
especially if there isn't a data scientist
on staff. And even if there is,
they might not have the time to constantly
monitor the program's data
for trends.
This is why hiring the right and
adequately trained personnel is emphasized
in point number three. If
the right team is in place, they will be
able to take on the task of analyzing
data and integrating it into a predictive-maintenance
program. Without
proper data analysis and integration,
a predictive-maintenance program
will be nothing more than a shot in
the dark.
Solutions. To ensure adequate data
analysis and integration, chemical
plant managers can consider the
following:
* Use data-visualization tools to
better understand the data and to
spot trends
* Automate data analysis as much as
possible, using tools like machine
learning and artificial intelligence
* Delegate data analysis tasks to
team members skilled in data
analysis and have the time to do
it properly
While there is more to data analysis
than what we've listed here, these
tips will get managers started on the
right track. Additionally,
it's important
to remember that data analysis
is an ongoing process - the more
data analysts have, the better their
insights will be.
5. Lack of consistency in predictive
maintenance implementation.
Consistency is an integral part of any
maintenance program, and predictive
maintenance is no exception. To
get the most accurate results from
a predictive maintenance program,
data must be collected consistently
and accurately.
This means having a plan in place
for collecting data, appointing who
will be responsible for managing it,
and what processes will be put in
place to ensure that data are collected
accurately. Without a consistent
approach to data collection, a
predictive maintenance program will
be less accurate and less effective.
Solutions. Here are a few tips on
how to maintain consistency in a
predictive maintenance program:
* Establish a straightforward proCHEMICAL
ENGINEERING WWW.CHEMENGONLINE.COM
MARCH 2023
http://WWW.CHEMENGONLINE.COM
Chemical Engineering March 2023

Table of Contents for the Digital Edition of Chemical Engineering March 2023
