Top of Form

Chemical

Chemical

Food & Beverage

Food & Beverage

Hydrocarbons & Energy

Hydrocarbons & Energy

Life Sciences

Life Sciences

LNG

LNG

Power

Power

Pulp & Paper

Pulp & Paper

Refining

Refining

Water & Wastewater

Water & Wastewater

DoctorKnow® Application Paper

Title: Intelligent Motors Moving to the Forefront of Predictive

Maintenance
Source/ Matt O'kane
Author:
Product: Motor
TechnologyMotor
:
Classificati Not Classified
on:
Copyright Material IEEE
Paper No. PCIC 2000-22
MATTHEW M. O'KANE
U.S. ELECTRICAL MOTORS
DIVISION OF EMERSON ELECTRIC
8100 W. FLORISSANT AVE.
ST. LOUIS, MISSOURI 631361
Abstract

This paper presents the technological advancements of embedded

sensors and microprocessors into ac induction motors and their ability to
predict failure. Today, maintenance departments utilize several stand-
alone technologies to detect and predict motor failure such as vibration and
flux analysis. Intelligent motors now integrate multiple technologies that
enable continuous monitoring of critical motor components as well as
operating conditions. In today's global economy, the need to increase
profitability while decreasing operating expenses sets the stage for the
emergence of Intelligent Motors.

Index Term - Motor.

INTRODUCTION

The new global economic reality of consolidation and downsizing has

significantly increased the awareness and the
impact of continuous machine operation to profitability. In the 1970's and
80's inefficiencies in manufacturing could be more easily hidden by price
increases attributed by inflation. Today, with all of the mergers and
acquisitions now occurring, the result is in more and more buying power
moving to fewer and fewer hands. Price increases in manufactured
products are not only unacceptable; they can also result in a significant
loss of business and customers.

The true measure of a company's operational asset effectiveness starts

with plant utilization. Significant process
improvements have been made in recent years with as an example, loop
optimization. However, the advantages gained
by this approach along with any other improvements can very quickly be
nullified by the unexpected failure of critical ac
induction motors. Additionally, many large end users are now focusing on
their core competencies which in many cased
does not include maintaining plant assets.
The theory of Reliability-Based-Maintenance (RBM) will be provided to
describe the overall goals and advantages to such
an approach. The objective is to show the relative importance and
feasibility to embedded sensors and microprocessors
within ac induction motors and their potential for enhancing overall plant
utilization and their corresponding positive economic impacts. Additionally,
the need to provide the end user with sensor data, which is easily
understood and can be
acted upon expeditiously, will also, be discussed. The seven parts of the
paper are as follows:
1. Part I - Root Mean Cause of AC
Induction Motor Failure
2. Part II - Theory of Reliability Based
Maintenance (RBM)
3. Part III - RBM Technologies
4. Part IV - Integrated Strategy for
Embedded Sensors
5. Part V - Diagnostic Software
6. Part VI - Hazardous Events
7. Part VII - Summary and Conclusion
PART I - ROOT MEAN CAUSE OF AC INDUCTION MOTOR FAILURE

Since the invention of the first ac induction motor, the reality of eventual
failure has never changed. What has changed
however is the understanding of why motors fail and efforts to improve
overall performance and life expectancy of the ac
induction motor. The forces that impact the dynamic performance of any ac
induction motor are many. The subject
of root mean causes of motor failure has been covered extensively in past
IEEE papers. Therefore Table 1 is intended
to provide a summary of the key components of the stresses that
contribute to ac motor failure.

PART II - THEORY OF RELIABILITY BASED MAINTENANCE (RBM)

To discuss Reliability Based Maintenance (RBM), we must first describe

what it is. RBM is a strategy to improve overall plant productivity by
integrating Preventative Maintenance (PM), which is time or interval based;
Predictive Maintenance (PDM) which is machinery conditioned based; and
Proactive Maintenance (PAM) which is root-cause/failure analysis based,
in a formal and balanced approach to modify the normal failure profile of ac
induction motors along with all other plant machinery.

Predictive Maintenance is the center point in this approach along with

possessing the tools to monitor existing driven
machinery condition. The result of this monitoring provides the necessary
information required to take pro-active steps once potential failure is
identified. However, to complete the RBM process, the need to determine
root cause analysis is
fundamental in achieving maximum available capacity of plant equipment.
Figure 1 illustrates the primary steps of the RBM process.

Figure 1: Reliability Based Maintenance

The type and kind of plant equipment that effects continuous production
ranges from valves and transmitters to pumps and
compressors. The ultimate goal of all maintenance activities is the
extension in the useful life-cycle of all machinery assets. When it comes to
ac induction motors, many companies have focused preventative and
predictive efforts on the largest horsepower motors. In many cases, the
size in motor horsepower is directly proportional to the criticality of the
process. The definition of critical would be if failure either shut downs or
limits production.

The traditional approach to the maintenance of ac induction motors has

been based on a financial culture, which places
greater weighting on the cost of repair, rather than the total cost to the
business interruption. Additionally, the majority of
ac motors which actually fall into the critical category are in the NEMA
sizes. However it is estimated that the NEMA size
motors are typically those which receive the least amount of RBM efforts.

The goals of implementing a comprehensive RBM strategy are numerous

and are summarized as:
1. Increased plant capacity/production
through the elimination of downtime.
2. Reliable knowledge of machinery
condition providing status of overall plant
capacity.
3. Extension of the useful service life of
plant machinery through identification and
elimination of failure modes at the source.
 4. Development of work teams
responsible for the combination of
maintenance and capacity.
5. Maintenance, Production, and
Engineering function as partners in
maximizing plant capacity.
6. Maintenance provides credible machine
information to assist in crucial decision-
making on plant operations.
7. Creation of a learning organization
where mistakes are not repeated.
8. Development of a systematic approach
for each maintenance situation (reactive,
preventive, predictive) based on logic and
history.
9. Continued and increased management
support for RBM strategies.
PART III - RBM TECHNOLOGIES

Multiple technologies exist today that are utilized in the area of predictive
and preventative maintenance. The most widely
implemented technology today is vibration analysis. However, with the
advancement in microprocessor technology, the result has lead to smaller
and less expensive portable equipment. Several companies provide
industry with an array of portable equipment capable of both measuring the
following:
1. Vibration analysis
2. Oil & wear particle analysis
3. Infrared thermography
4. Precision alignment and balancing
5. Motor stator flux
6. Motor current signature analysis
7. Ultrasonics

The implementation of these multiple technologies provides a method for

the user to identify potential problems with the ac
induction motor along with other rotating equipment. A shortcoming of
several key technologies such as vibration is
that they will view data for the short instant in time while they are
connected to the motor. The result of this approach is that it does not
provide for any operational trending information or synchronization of the
data points in time.

The fact is many failures are caused not by the failure of the components
due to design or poor quality but rather by operational anomalies. Poorly
trained personnel along with the need to push machines beyond their
design characteristics for increased production also contribute to the failure
cycle. In many cases, temperature and lubrication play
significant roles in the failure of ac induction motors.

The move to integrated embedded sensors within the motor provides the
ability to trend and view the operational and
functional characteristics real time. The ability to embed several sensors
that monitor key functional parameters
coordinated with a dedicated microprocessor provides unparalleled
monitoring, data fusing and predictive information
enabling the user to decrease the risk of unscheduled downtime.
PART IV - INTEGRATED STRATEGY FOR EMBEDDED SENSORS

In most process, manufacturing and comfort applications the sudden and

unexpected failure of ac induction motors can
cause significant disruption leading to loss of productivity as well as
profitability. As we have discussed, Reliability Based
Maintenance strategies and technologies aid in the detection and
prevention of unexpected motor failure. However these
current methods require sophisticated instruments and highly trained
personnel to implement this approach.

In addition, most maintenance departments cannot monitor all of the

motors within a given plant or facility. The cost in labor hours required to
complete and maintain this task, would for most companies, be financially
impossible. The combined decrease in cost and increase in the
sophistication of sensors, intelligent algorithms, processing and memory
power now makes the use of intelligent motors in industrial processes
desirable.

The placement of critical sensors to monitor vibration, flux, insulation and

temperature permits implementation of data
analysis techniques that result with the prediction of motor life. This
approach will complement most current RBM strategies along with
significantly increasing the coverage of ac induction motor monitoring to a
horsepower level not currently covered by most route based maintenance
strategies.

The initial version of the intelligent motor will apply only to fixed speed
sinewave ac induction motors. Variable frequency drive compatibility will
follow in the next generation of intelligent motors.
SYSTEM OVERVIEW

The heart of the intelligent motor is a 16-bit microprocessor with 14 bit A/D,
1 Mbytes RAM memory and 0.5 Mbytes of
Flash Memory. The electrical components on the printed circuit boards are
industrial grade and will withstand temperatures in excess of 85 degrees
Celsius. Communication is via RS485 and the output data is viewed in a
very user-friendly software environment. The specific nameplate motor
data and additional design characteristics per each motor is loaded into the
memory during the manufacturing process. See Figure 3 and Table 2
which describes both the system architecture and senor allocation.
 Figure 3: System Architecture

The intelligent motor parameter values along with flux and vibration
spectral data are processed and trended based on
the following measurement categories combined with fundamental alarm
and advisory logic:

Operational measurements are computed and trended through a

combination of the vibration and flux sensors. Significant
parameters that are measured and trended include:
1. RPM
2. Load
3. Number of Starts
4. Run Time

RPM is calculated through vibration and flux measurements and is

accurate within +/- 1 RPM. Motor load is then calculated based on slip as
the motor's rated RPM. Motor starts are detected and time stamped which
provides for knowing the exact operational condition of the motor. Motor
run time is also calculated and provided which assists in obtaining Mean
Time to Repair (MTTR).
Electrical components are monitored and trended by measuring the
motor's electromagnetic field with an internal flux coil mounted axially in the
motors end bell. High resolution Fast Fourier Transform (FFT) spectrum
analysis of the flux field results in a flux signature containing valuable
information related to electrical components of the motor including:
1. Voltage Phase Imbalance
2. Stator Turn to Turn Shorts
3. Rotor Bar Condition
4. Ground Fault Condition

Voltage imbalance as low as 3% would require that the motor be de-rated

to approximately 85% of design load rating or
suffer the consequences of premature motor failure. A severity indication
related to any of these values is determined through comparison with
detailed measurements of motor current spectrum of flux at full load.

High motor vibration is another major contributor that will shorten the life of
the motor as well as cause catastrophic
failure. Mechanical components are monitored and trended by measuring
the motor vibration through a radial vibration
transducer. High resolution FFT spectrum analysis (up to 2500 Hz and
3200 lines) of the vibration signature will provide
trend information related to:
1. Balance
2. Alignment
3. Bearing Condition
4. Belting Issues
5. General System Problems

One of the most under-used, yet useful measurements available for

proactively and preventively monitoring motors is
temperature. The intelligent motor monitors bearing temperature at both
bearings to primarily assist in predicting and preventing bearing failures.
Another temperature probe is embedded in each stator winding to provide
valuable information as to the condition of the stator. By combining a frame
temperature and ambient temperature along with various combinations of
the other temperatures, parameters are calculated (see Table 3) to monitor
the operating condition of the motor. The intelligent motor incorporates 3
patented insulation sensors (1 per phase) which are embedded in the first
winding turn during the winding process and will alert the user once
approximately 50% insulation life has passed.

Additionally, the intelligent motor incorporates two 'Form C' contacts.

These two contacts will close when level 1
& level 2 hazardous events have occurred. This provides a direct
connection which can alert the user through a Programmable Logic
Controller, Distributed Control System or simply lighting a light in the
maintenance office. This provides the user with immediate notification to
any serious condition prior to failure.
* Parameter to be added at a future date
Note: LF = line frequency
NP = number of poles
SF = slip frequency (Synchronous speed (Hz) - Running speed)
n = 1,2,3,…
PART V - DIAGNOSTIC SOFTWARE

Embedded sensors do not automatically equate to better information. The

key to utilizing the sensors is to develop user
friendly and concise diagnostic software. Today's technology in many
cases provides too much information for the end user. The information
most users prefer is that which is provided in a summarized manner along
with corresponding
recommendations. This is accomplished in four steps:
1. Simplistic summation of data retrieved
from all sensors
2. Viewing of multiple attributes
simultaneously
3. Operational trending data
4. Diagnostic software recommendations
developed by experience.
 Figure 4: Motor Condition Analysis Report

The summation page shown in Figure 4 displays in an easy-to-read format,

each attribute gained from the embedded sensors is displayed with an
individual alarm indication. The key is to have a very quick review of the
motor's overall health and to see what potential problems may have arisen.
This summation minimizes the time for first review and allows
maintenance to focus on problem motors, not conducting exhaustive
diagnostics on healthy ones. The individual
attributes alarm thresholds are presented with historical data and may be
changed by the user to meet their particular
application or requirements.

As seen in the summation page shown in Figure 4. The condition of

multiple attributes are displayed such as load,
ambient, frame temperature and vibration. Figure 5. Shows the specific
middle synchronous vibration trend over the trend
period. Similar trends are available for all attributes shown such as
vibration, flux, temperature and load. This illustrates
the trending information available and the effects which provides a view to
the functional as well as operational conditions of the motor. Using the
advancements in Windows95® software, multiple trends can be viewed
simultaneously which provides for a comprehensive review of the critical
elements effecting the motor.
An example of an actual motor within a pulp mill in North Carolina that as
been fitted with on board sensors is shown.
The trend data for the acquisition interval illustrates excessive vibration
with the 60HP pulper motor. This particular motor was found to have an
alignment problem caused by the coupling. The motor was aligned, the
coupling greased and the pump packed. Once the maintenance was
completed, the middle synchronous vibration decreased.

Figure 5: Average and maximum values of the middle

synchronous (3 to 8 x RPM) vibration energy
for a 60 hp pulper pump motor.

Figure 6 illustrates the Load Profile of the same 60 hp pulper motor and the
random load spikes were determined to be
caused by process problems.

Figure 6: Average and maximum percent load values for a 60 hp motor

driving a pulper pump.

The final example (Figure 7) of the 60 hp pulper motor is the actual heating
parameter during the same sampling period. The motor was determined to
have a clogged fan cover which caused the motor temperature to increase
by over 11 degrees C. Once cleaned, the motor returned to a normal
operating temperature.
 Figure 7: Average and maximum values of the heating parameter
is the delta between the ambient and frame temperature
for the 60 hp motor driving a pulper pump.

PART VI - HAZARDOUS EVENTS

One of the most important features of the intelligent motor is the ability to
detect and warn of hazardous events. A
hazardous event is defined as any extreme operating characteristic
exhibited by the motor that indicates imminent
failure. Hazardous events are categorized as either a level one or level two
alarm. Level one alarms are triggered for the more serious and urgent
occurrences. Definition of the hazardous events to be monitored and the
annunciation levels for each are listed in Table 4.
Because the spectral and trend data immediately associated with a
hazardous event are stored, (as seen in Figures 5, 6 & 7) the best
information possible is available for analysis. With this data, one can
potentially identify the root mean cause of the hazardous event. The data
may also include early indications of impending motor failure or a record of
improper motor use, which will lead to mechanical and or electrical failure.

PART VII - SUMMARY AND CONCLUSIONS

In today's global economy, the need to maintain production availability

along with the pressures of decreased overhead
while increasing profitability is changing the way maintenance is
performed. We have discussed the multitude of stand alone external
technologies which aid in the detection of motor failure. However, with the
advent of intelligent motors, the
benefits will far outweigh the initial capital costs by more effectively utilizing
route based maintenance labor and most importantly, preventing
unscheduled downtime.

Finally, the technology that is now being applied into motors is just the first
step in the proliferation of similar technology into other critical driven
equipment such gear box's, compressors, pumps and fans. As the
adoption rate increases, so to will the volumes, which will in turn reduce
the cost. As seen with other microprocessor based products, the reduction
in price will allow the integration of intelligent sensors into the lowest
horsepower rating of motors and other critical components. The final
outcome will be further production capacity increases and most importantly
increased revenues for those companies who embrace and implement
Reliability Based Maintenance strategies inclusive of the "Intelligent
Motor".
REFERENCES

[1.] Austin H. Bonnett, "A Methodology for Root Cause A.C. Motor Failure
Analysis".
[2.] David G. Petersen, "The Business of Plant Maintenance", October
1995.
[3.] Paul A. Wishnafski, "The Road to RBM: Implementation of a Reliability-
Based Maintenance Program", Computational Systems Incorporated,
1997.
[4.] Stewart V. Bowers, III, PhD., "Implementing the Status Smart Sensor
Technology into a Predictive and Proactive Maintenance Program",
Computational Systems Inc.1997.
[5.] Stewart V. Bowers, III, PhD., "Intelligent Motors in Industrial Settings",
Computational Systems Inc. 1999.
[6.] Mike Klein, "The Next Generation of Manufacturing Automation, It's All
About Productivity". 1999.
VITA

Matthew O'Kane is Director of Synergy Projects for the Industrial Motors &
Drives business unit of Emerson Electric
Company. Matthew is responsible for the development and implementation
of embedded motor technology between the
U.S. Electrical Motors, Leroy Somer and Computational Systems, Inc.
Divisions of the Emerson Electric Company. This
includes directing all corresponding commercial and technical activities
with embedded motor technology. Matthew has spent most of his career
 working with industrial automation products and services for both Reliance
Electric and Emerson Electric. Matthew received his BS in Electrical
Engineering Technology Degree from the University of Maine.

All contents copyright © 1998 - 2006, Computational Systems, Inc.

All Rights Reserved.

Send comments to: Last Updated 05/08/03 © Emerson, 1996-2005

mhm.info@compsys.com Legal and Privacy Statements