Chemical Engineering February 2023 - 40

Problem 2:
Upskilling
personnel
FIGURE 2. Reading waveform data, such as the spectrum shown here, can be
a complex task for inexperienced personnel
such as a tablet or mobile phone.
This untethers digital workers from
manual rounds, and even from the
control room.
Instead of collecting data every
couple of weeks or months, wireless
sensing technologies enable
plants to collect critical data every
few minutes or hours. Reliability
teams can maintain constant vigilance
over assets, regardless of
their criticality, helping to eliminate
unplanned outages and unexpected
emissions for more reliable, sustainable
operations.
Small sensors bring big results
For one large North American power
producer,
small wireless
sensors
installed on a cooling water pump
avoided a big hassle.
The plant had moved from manual
maintenance rounds to wireless
vibration monitors, which delivered
spectrum, waveform, and embedded
analytics to maintenance personnel
and operators in the control
room. Shortly after installation of
the wireless sensors, maintenance
personnel were notified of a critical
alert status regarding lubrication of
a cooling-water pump motor.
The
waveform
data
indicated
there was a problem with impacting,
and a lubrication display in the
red zone showed the cooling-water
pump likely had a dry motor bearing.
Technicians greased the bearing and
collected vibration data on demand,
this time with the impacting eliminated
and lubrication status in the
green zone due to dramatically lower
vibration levels.
By setting up their reliability equipment
to deliver actionable information
to the right people at the right
time, the plant avoided a potentially
costly equipment failure and extended
downtime.
40
One of the key metrics
by which reliability
teams judge
their success is response
time. When
there is a problem
in the plant, teams
should be able to
respond quickly
and effectively -
a task that was much easier ten or
fifteen years ago, when plants often
had a " deep bench " of expert personnel
to identify, analyze and solve
equipment issues.
Today, however, plants are running
with far fewer personnel, and those
people often have far less experience.
Most vibration-related equipment
problems will fall into one of a
few common categories:
* Balance
* Alignment
* Looseness
* Blade pass
* Bearing
* Lubrication
However, identifying those problems
from just raw spectrum and
waveform data can be a complex
undertaking (Figure 2).
A new technician typically needs
two to three years of experience before
he or she can analyze raw data
and identify the root cause of a problem.
Waveform data identifies the
severity of a problem, but without an
identified source, that information is
difficult to use.
Severity and source are interlinked,
and manually making that connection
using spectrum and waveform
data to identify the most common
equipment problems will be difficult
for new personnel. Waiting two to
three years for them to gain these
skills has the potential to result in
significant downtime and loss.
Solution: Decision support
Instead of waiting for new technicians
to come up to speed over several
years, today's reliability teams are implementing
decision support through
edge-analytics software to more
rapidly upskill and assist personnel.
Many of today's wireless edge devices
contain embedded prescriptive
analytics tools, effectively building expert
knowledge into the device.
Instead of spending hours evaluating
and manually analyzing raw data,
digital workers simply access the report
provided by the software on the
edge device. In the most advanced
edge devices, these tools are intuitive
enough to be used by a person
with no analytics experience whatsoever.
For example, equipment health
can be identified by color - green,
yellow or red, based on severity of
defects - and the software can inherently
identify the most common
equipment faults (Figure 3).
Using the intuitive screens instantly
displayed on their console or web
device, personnel can address problems
immediately (when possible),
flag them for maintenance during a
scheduled outage, or travel to the
asset with a handheld analyzer to
gather more information for deeper
troubleshooting. Traveling to the machine
with an analyzer is the exception,
not the standard, ensuring the
time spent manually investigating
equipment only occurs on the small
number of failing machines.
Routes-by-exception
One North American manufacturer
was producing products that
became very popular during the
COVID-19 pandemic, forcing them
to rapidly increase production. Upping
production increased the potential
for
problems,
and it
also
meant less room for error because
any downtime caused by equipment
outages would result in lost production
of critical supplies.
An edge device with embedded
analytics
software
detected
and
identified lubrication issues on six
assets, enabling the maintenance
team to intervene before failure.
Moreover, it also triggered vibration
alarms as supply fans began experiencing
problems during production
increases. Immediately following
up on the vibration alarm not only
helped the maintenance team idenFIGURE
3. Analytics tools for machinery analysis,
provide a simple and reliable indication of
equipment health
CHEMICAL ENGINEERING WWW.CHEMENGONLINE.COM
FEBRUARY 2023
http://WWW.CHEMENGONLINE.COM
Chemical Engineering February 2023

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