PDM in Action-eBook
PDM in Action-eBook
com
ment reliability
s. LUDECA is your
er 30 years of
ent to quality, we will
PdM
IN ACTION
t running.
PdM Assessment
P.2 [CLICK HERE]
Foward Focus
P.14 [CLICK HERE]
How is PdM being implemented Group, at Allied Reliability Group PM program. We are seeing a few
in the average plant today? In a joint (www.alliedreliability.com). “Once, more people begin to treat it as im-
survey conducted by Plant Services it was this electronic trickery and portant, and sometimes even more
and ARC Advisory Group (www.arc- considered less important than the so than the PM program.”
web.com), maintenance and reliability
professionals shared their experiences
and insights into the day-to-day use of WHAT’S ON TAP
PdM on the plant floor.
Many plants are ahead of the curve
and already have some form of predic-
ELECTRIC MOTOR TESTING
PREDICTIVE MODELING
CORROSION
SOFTWARE
VIBRATION
INFRARED
ACOUSTIC
n Satisfactory
REDUCE OPERATIONAL COSTS
n Effective
n Very effective
IMPROVE UPTIME
Figure 2. When asked to rate the performance of their PdM program over the past 12 KNOWLEDGE CAPTURE
months, more than half admitted their programs need some improvement.
INDUSTRIAL-GRADE TABLET
CONSUMER-GRADE TABLET
INDUSTRIAL-GRADE PC
PAPER-BASED SYSTEM
CONSUMER-GRADE PC
NO IMPROVEMENTS NEEDED
INDUSTRIAL-GRADE
CONSUMER-GRADE
SMARTPHONE
SMARTPHONE
OTHER
BUDGET CONSTRAINTS
UNDEFINED FINANCIAL
BENEFITS
LIMITED IT RESOURCES
Figure 5. When asked which data collection methods they were using with their PdM
systems, a surprising number of respondents admitted to using a paper-based system,
with industrial- and consumer-grade PCs rounding out the top three.
IN-HOUSE OPERATIONS
IN-HOUSE RELIABILITY
OEM SUPPLIER
FROM EQUIPMENT SUPPLIERS
ENGINEERS
MONITORING
OTHER
WIRELESS CONNECTIVITY TO
FIELD WORKFORCE
(WORK HISTORY, MANUALS, ETC.)
SMARTPHONES
some of their work is done in air-con- asked which emerging technologies
ditioned spaces in front of computers. they were considering for inclusion in
Also, some of the conditions they their PdM initiatives. Within the next
TABLETS report are not verifiable with the five three years, many respondents plan
senses of repair personnel and they to have analytics software, tablets,
distrust the finds reported because the wireless connectivity to workers in the
asset appears to be working perfectly field, and embedded PdM intelligence
RFID FOR ASSET IDENTIFICATION well when the recommendation is from equipment suppliers (Figure 6).
made to fix something that doesn’t However, it looks like GIS for asset
appear to be broken.” location, RFID for asset identification,
In this age of smartphones and and cloud-based PdM are still out of
GIS FOR ASSET LOCATION bring-your-own-device (BYOD), it’s reach for many plant workers.
easy to imagine an interconnected “I think we are on the edge of a sub-
plant where machines talk to one stantial change in the PdM technolo-
another and diagnose problems on gies,” says Isenhour. “Based on im-
ANALYTICS SOFTWARE their own. But is this type of techno- provements in technology and current
logical communication happening in trends in the industry, I believe that
most plants, or is it still just a dream? we are about to see a consolidation of
When asked which data collection the PdM technologies. Picture an iPad
methods they were using with their that could operate as an infrared cam-
OTHER
PdM systems, a surprising number of era, a vibration analyzer, ultrasonic
respondents admitted to using a pa- analyzer and CMMS reporting device.
per-based system, with industrial- and One box covering three or more tech-
consumer-grade PCs rounding out nologies and combining them to allow
n Have now n In 2014 budget
the top three (Figure 5). Most survey for asset health instead of faults by
n Within three years n No plans
participants had no plans to imple- technology. We are very close. We can
ment smartphone or tablet technology already access our CMMS from many
Figure 6. Within the next three years,
many respondents plan to have analytics
in the near future. handheld devices.” Vendors have cre-
software, tablets, wireless connectivity to Although most respondents seem ated infrared modules that work with
workers in the field, and embedded PdM to be taking a more traditional, less the iPhone, and vibration is available
intelligence from equipment suppliers. technological approach to PdM, we for the iPad, he explains.
SYSTEM INTEGRATION
LABORATORY INSTRUMENTS
RELIABILITY SOLUTIONS
EAM/CMMS SYSTEM
EH&S SOLUTION
OTHER
ERP SYSTEM
HISTORIAN
OTHER
50
25
0
Plant Maintenance Maintenance Maintenance Reliability Reliability Controls All
manager manager engineer technician engineer technician engineer others
WHICH ARE YOU USING CONTROL SYSTEMS AND CONNECTED SENSORS TO MONITOR?
50
25
0
Automation Control Production Distribution Electrical Fleet HVAC/R Safety Manufacturing Other
assets system assets pipelines systems vehicles system systems/ productivity
assets devices
Self-diagnostic (smart) instrumentation and final control elements (drives, valves) 51.4% 5.1% 13.0% 30.5%
Specialized system control systems (combustion controls, steam trap monitors, water/
52.5% 5.1% 10.7% 31.6%
wastewater, cooling systems, compressed air)
Predictive maintenance (PdM) technologies for condition-monitoring (thermography, oil
70.1% 4.5% 8.5% 16.9%
analysis, vibration, ultrasound, motor testing)
Instrumentation (valve analysis, signal processing and alarm management software),
52.5% 4.5% 15.8% 27.1%
automation, data archiving, and reporting software
Automation of work order generation (including repair history, diagnostics, procedure
48.6% 15.3% 18.1% 18.1%
and parts information)
Energy monitoring and control devices (drives, valve actuators) 41.2% 9.6% 16.9% 32.2%
Tranter. “Lack of executive support and budget constraints added that he believes the low share of “satisfactory” rat-
often flow from a lack of understanding or belief in the ings “relates to the fact that most PdM technicians are not
financial benefits of condition-based maintenance. That adequately trained on the technologies, and when they do
situation often exists because there is a lack of agreement on provide recommendations, they are often too technical and
how to quantify the benefits of detecting a failure before it vague.” Most users of PdM tools need “clear, unambiguous,
occurs. This is one reason why benchmarking and tracking actionable information,” he says.
KPIs is so important, along with active communication.”
Timothy Dunton, a director and instructor and developer
Few readers anticipate seeing an
of Reliable Manufacturing at Reliability Solutions Training
(www.reliabilitysolutions.net), says that old-school thinking increase in PdM investments greater
at smaller plants may be stalling more-widespread adoption than 5%, but of more importance,
of PdM tools and practices. “You can get into the game quite very few anticipate trimming
inexpensively with the right approach,” Dunton says. “It
would seem that we have done a poor job of demonstrating
investments in their PdM programs.
the value of lower-cost tools.”
But without adequate resources (whether internal or In addition, Tranter says, if there’s not widespread buy-in
external) to help implement PdM tools, those advocating within an organization into the philosophy of condition-
at smaller plants for adoption of PdM strategies may find based maintenance, then reactive or preventative mainte-
themselves stuck between a rock and a hard place. Azima nance strategies will dominate.
DLI’s Hurlock comments, “What right-minded organization Reliability Solutions Training’s Dunton sees reason for
invests in anything without defining financial and operat- predictive maintenance champions to take heart. Among
ing benefits, and without executive support and the IT and plants that have begun to incorporate PdM tools, “there is
engineering resources required to sustain the program? ... an awakening that (PdM programs) have more to offer,”
If benefits and objectives can’t be defined and resources and he says. That’s one way to look at the large share of survey
support aren’t available, PdM is a waste of time and money.” respondents saying that their plant’s PdM program could
Survey respondents clearly see room for improvement in use improvement, he offers: They recognize that PdM has
their plants’ own PdM programs. Forty percent of those whose potential that’s not yet being realized.
facilities have deployed PdM tools said that their PdM pro- “For that to happen,” he says, “analysts will need more
grams need improvement. Only 25% rated their plant’s PdM time, and for that to happen either the (user) groups will
program as satisfactory. have to grow, or we leverage technology to eliminate the
“It is unfortunate to see that only 4% of people feel more mundane aspects of the process. Lower-cost tech-
that their PdM program is ‘very effective,’ ” says Mobius nology, simpler technology, might enable more people to
Institute’s Tranter, reviewing the survey results. Tranter be involved.”
Figure 1. Some
Defect initiated Potential failure technologies
Ultrasound emitted
provide earlier
Vibration detectable warnings than
others.
CONDITION
Particles detectable
J F M A M J J A S O N D J F A
MONTH
Source: Fluke
Figure 2. Regular, periodic vibration testing and analysis has been Figure 3. Thermography provides great detail in reasonably short
shown to be the most broadly applicable PdM technology. time periods without necessitating physical contact with equipment.
“Criticality of the assets to the production output and worst actors – the most expensive to maintain and least reli-
safety of the facility should play a big part in the initial PdM able – because if you don’t get early successes, the program
application,” suggests Shon Isenhour, a partner at Eruditio will lose favor when the novelty wears off,” advises Timo-
(www.EruditioLLC.com). thy Dunton, director, instructor, and developer of Reliable
“A criticality analysis ranks the assets based on the risk Manufacturing at Reliability Solutions Training (www.
they present to your company in the areas of safety, health, reliabilitysolutions.net).
environment, business impact, quality, and other factors,” “Apply PdM to the top 20 percent of the critical equip-
explains Bill Barto, director of the reliability solutions ment list first, and after the health of those assets improves,
group at Life Cycle Engineering (LCE) (www.lce.com). “For tackle the next 20 percent, and so on until more than 85
example, I would target a pump whose failure causes an percent of your assets are covered,” suggests Andy Page,
immediate process disruption (high business impact) over a principal consultant for the technical services group at Al-
pump that has a backup on standby (low business impact).” lied Reliability Group (www.alliedreliabilitygroup.com).
Jason Tranter, founder of Mobius Institute (www.mobi-
usinstitute.com), takes that concept further. “Without an MAKING THE MOST OF YOUR BUDGET
understanding of the failures modes, including an appre- PdM funding commitments may initially be difficult to
ciation of the P-F interval (the time from first symptom to win, so choose an approach that delivers substantive wins in
functional failure), it is not possible to determine which order to justify continuing investments.
PdM technology can detect the fault or how frequently tests “Businesses tend to make emotional decisions regarding
must be performed,” says Tranter. how to manage PdM,” says Reliability Solutions’ Dunton.
“We look for low-hanging fruit as well as the top 40 list of “If they have a failure that they blame on a lack of vibra-
Source: UE Systems
as compared to their acquisition and
training costs.”
He encourages leveraging existing
investments. “If you already have vi-
bration analysis technologies available
at your site, you would target critical
rotating assets. If you already have
ultrasonic equipment, you might focus
on compressed air leaks and ultrasonic
lubrication,” explains Isenhour. Figure 4. Ultrasound can offer one of the fastest ROIs of any PdM technology.
“I would start with the PdM tech-
nology that everyone already pos-
sesses: their human senses,” says LCE’s the organization raphy services. “Vibration analysis
Barto. “Begin with simple visual and “Infrared thermography covers enables earlier detection and diagnoses
audible indicators of how well your all asset types, is relatively easy to problems, although it requires direct
equipment is running, and resist the learn, is much cheaper than vibration contact with equipment. Motor testing
temptation to run out and buy other analysis, and it’s very easy for people is one of the most common forms of
equipment before getting used to the to accept the information coming reactive testing, but it can be incorpo-
idea of basic condition monitoring. from the camera as the truth,” says rated into a more proactive program.”
“A good example is a client that Page. “The same can be said for pas- “Regular, periodic vibration test-
had many air conditioning filters that sive ultrasound, but it is not the case ing and analysis has been shown to
were replaced after a certain number for vibration analysis.” be the most broadly applicable PdM
of days. I recommended a device that “No single technology is better technology,” suggests Joe Van Dyke,
used the differential pressure across than the other,” remarks Tyler Evans, VP of operations at Azima DLI (www.
the filter to ‘whistle’ when it became business unit manager of vibration azimadli.com).
dirty. Now, they listen for the whistle and alignment products at Fluke “Ultrasound can offer one of the
and replace the filter only when it is Corp. (www.fluke.com). “A plant with fastest ROIs of any of the PdM tech-
sufficiently dirty,” explains Barto. lots of motors may want to focus on nologies,” says Doug Waetjen, VP
vibration and thermography. One of global operations at UE Systems
HOW SHOULD TECHNOLOGY with many gearboxes and drive shafts (www.uesystems.com). “It can be used
CHOICES BE MADE? might benefit more from oil analysis on such a diverse number of assets
New technology investments should and thermography. A plant with a lot and failure modes for compressed
be based on three criteria, according of automation and electrical control air leaks, electrical inspection of
to Allied Reliability Group’s Page: systems may lean toward thermogra- energized assets, bearing testing and
1. Effectiveness of the PdM technology phy and power quality tools.” trending, and condition-based lubrica-
across all asset types “Thermography offers great amounts tion. Ultrasound also has one of the
2. E ase of implementation (e.g., for of detail in reasonably short periods of shortest learning curves.”
learning the hardware, software, time without physical contact with the One successful approach noted by
and technique) equipment,” explains Michael Stuart, Waetjen is to use ultrasound to conduct
3. Degree of reception from those in Fluke’s senior manager of thermog- an audit of your compressed air system,
One of the factors separating process plants that operate reasonably well from those
that achieve excellence is how they use information. Best-performing plants achieve
their status by reducing production outages through proactive maintenance. This in-
creases revenue generation and lowers maintenance costs to drive profitability higher.
Some companies enjoy this state of affairs as a way of life, while others ask,
“How do they do it?” A key difference is the ability to use instrument diagnostic
information effectively, but how is such a concept put into practice?
Plant personnel have dashboards showing how various units within a plant are
performing; they can see how a reactor, a heat exchanger, a boiler or some other
major piece of equipment is performing at any time. For each of those units to
perform optimally, control loops supporting them must perform as designed, and
that depends on instruments operating reliably to provide accurate data.
Smart instruments – including transmitters, sensors, and actuators – are capa-
ble of monitoring their own performance, and they can tell you a great deal about
what’s going on with a process well beyond the specific variable they measure or
the function they control. Just as the control system depends on these smart in-
struments, so, too, can your production and maintenance teams depend on them.
Some smart instruments are able to detect deterioration in a key component or rec-
ognize sensor drift. Others look at the process and might spot changes. For example, a
differential pressure sensor can measure process noise and determine the normal level
(Figure 1). If the noise level begins to fall or if differences emerge between high and
low readings, there may be a problem with impulse line plugging. Such plugging can
mask the true pressure differential and result in an unreliable reading. Today’s smart
instrument can send a warning when it begins to notice such changes.
Orifice
Blockage at Blockage at
No Blockage Pipeline
High Side Low Side
DP (Differential
Pressure)
PH (Static
Pressure H Side)
PL (Static
Pressure L Side)
3-valve manifold
CHANGES IN PROCESS NOISE CAN INDICATE PLUGGED IMPULSE LINES
Figure 1. By monitoring process noise, a pressure sensor can determine
whether impulse lines are plugging. Differences between the high side and
low side suggest one of the lines is becoming obstructed.
DP Transmitter
Level 2 comprises all the small groups of individual from smart instruments can help you identify root causes so
loops working together to provide basic functions. Those they can be fixed. Level 1 device diagnostics provide visibility
functions work properly when all the individual loops are into assets so you can see the problem clearly.
performing per design. With enough information from Once the bottleneck in an asset has been removed, the
smart instruments, plants can establish performance levels asset can perform with a higher level of reliability and ef-
for those groups of loops and subunits. ficiency, which will improve the whole plant. Once one bad-
Level 3 is the operational level. These are the asset actor problem is solved, the next step is to attack the next
diagnostics that operators and plant managers watch on asset that comes to the top of the list. One by one, problems
the main dashboards; among them are heat exchangers, can be identified and eliminated by using a well-thought-out
reactors, mixers, and the like. These assets are dumb in and strategy of analysis based on diagnostic information.
of themselves, so the only way to determine how they’re Having accurate performance information and eliminat-
performing is to go back down the chain and see what’s hap- ing bad actors is certainly valuable, but it provides only part
pening at Level 1. Smart instruments are the eyes and legs of the picture. Most companies find even greater payoff from
of this process, and information they provide is the basis for using diagnostics as the basis for a comprehensive predictive
everything higher up the chain. maintenance program.
You can’t do unit diagnostics without loop diagnostics, Most maintenance programs begin as reactive: Run
and you can’t do loop diagnostics without device diagnos- equipment until it fails and then fix it. This simple approach
tics. Operational effectiveness always finds its way back to isn’t optimal, but countless companies use it, as many assets
individual field instruments, sensors, and actuators. can be run to failure without serious consequences. A big
drawback of this approach when applied across an entire fa-
SOLVING PAIN POINTS cility is that it causes outages and production interruptions.
Most plants have trouble areas that continually cause prob- At a slightly higher level of sophistication is scheduled
lems. These bad actors cause an inordinately large number of maintenance. This is generally an improvement over a purely
outages and soak up more than their share of maintenance reactive strategy, but it is still an expensive method because a
dollars. Those two factors taken together rob the plant of out- device scheduled to be replaced may be functioning just fine
put and profitability. Going through maintenance orders and and does not need replacement. Some studies suggest that 12
reports typically makes them easy to spot and categorize. percent of maintenance costs are wasted because maintenance
Once these trouble spots are identified, the next step is de- is unneeded, with a given function performed simply because
termining the root cause of the problem. Does the same pump it’s scheduled.
drive motor burnout again and again? Does the same valve If reactive and scheduled approaches are the basis for
stick open? Is the energy efficiency of one heat exchanger far your maintenance program, you will have a difficult time
worse than that of others like it? The information available reaching operational excellence. But there is another way:
4
layer of a plant depends on
to happen, you can take action before
information from the one
nce
failure. below it. Supporting the maintena
Optimumoperation
What has to happen to make such a pyramid is the diagnostic &
capability an everyday experience? An data available from
3
effective maintenance program based smart instruments.
TIaO NS ostics
on condition monitoring has to have
three elements: OsPsetEpRerfAorm nce diagn
ency
A men t effici
• Field devices must be “smart,” mean- for equip
2
ing they have diagnostic functional-
1
• The main process control platform
T
he electronics built into smart instruments have be- into the flow. They can then intuit that the flowmeter
come more sophisticated over the past few decades. should be backflushed at the earliest opportunity,
Suppliers added self-monitoring capabilities to warn allowing the problem to be addressed in a way that
if something was happening with a smart instrument that minimizes disruption.
could impair its ability to deliver a reliable reading. While When maintenance operations are performed only
highly valuable, those functions were purely reactive and where they are needed, unnecessary work is reduced
could respond only after something had already happened. or even eliminated. In day-to-day operation using a
predictive maintenance program, the PAMS provides
Over time, those capabili- processing to take raw data the central evaluation and scheduling functions.
ties became more sophis- from the transducer (strain It takes diagnostic data from smart instruments
ticated and were able to gage, capacitive, magnetic, and combines it with heuristic diagnostics from
predict when a problem such etc.) and convert it to a plant personnel. The PAMS includes a mechanism
as impending component linearized signal within a to evaluate human-provided information as well as
failure or sensor drift was suitable range. Many smart electronic data, as smart operators can often detect
developing. Other diagnostic instruments required a problems better than any software platform.
functions allowed a smart secondary measurement The PAMS takes all the information and processes
instrument to moni- to correct the first, such it, perhaps drawing conclusions from multiple events
tor and record the as a temperature read- that might not be problems individually, but when
amount of force ing to compensate output taken together could signal trouble. It then sends
necessary to close from a strain gage. information to the right people and departments,
a valve or the As these smart instru- with each type of information getting to the people
amount of time ments became more so- responsible for acting on it, which drives the following
spent in any phisticated, suppliers added improvements:
given posi- internal monitoring functions • Operators can change some element of the process
tion. Soon it to alert operators of voltage if a smart instrument is exhibiting a problem and it
became clear problems, failed compo- can’t be fixed until the next shutdown
that these nents, excessive temperature • Maintenance scheduling can identify high priority
functions also excursions, and so forth. tasks
could offer in- Maintenance technicians • A purchase order can be issued for a valve seat
sights into the trying to solve a problem because a valve is beginning to lose its seal and will
process itself could use this information need to be rebuilt within the next two weeks
beyond the basic to verify that an instrument All of these elements combine to create an op-
process variable. or actuator was function- timized predictive maintenance program where a
The develop- ing properly or to trace process can run with maximum availability and
ment of communication the source of a problem. stability. Interruptions are minimized, as are overall
protocols such as HART, At first, these monitor- maintenance costs. Stores of replacement parts can
EtherNet/IP, Foundation ing functions were strictly be reduced as the number of repairs goes down and
Fieldbus and Profibus PA reactive, but with growing as they can be done with greater predictability. Such
provided a mechanism for sophistication made pos- an optimized predictive maintenance program is a
smart instruments to send sible through less-expensive critical step to operational excellence, and it all begins
information to the con- electronics, it became a with smart instruments and device diagnostics.
trol or asset management simple matter to add new
system, avoiding the need capabilities, including some Amit Ajmeri is an asset management solutions consultant
to visit a device in the field predictive in nature. Now, at Yokogawa Corp. of America. Prior to joining Yokogawa, he
to read and gather data. these advanced functions spent 12 years with Emerson Process
With the development of have become standard Management promoting Foundation
smart instruments, suppli- across virtually all smart Fieldbus technology and interoper-
ers added electronic signal instrument suppliers. ability. Contact him at amit.ajmeri@
us.yokogawa.com.