...

IN THIS ISSUE:

by user

on
2

views

Report

Comments

Transcript

IN THIS ISSUE:
Iris Power Engineering
Your Source For Monitoring the Reliability of Electrical Equipment
SUMMER 2004
IN THIS ISSUE:
PARTIAL DISCHARGE TESTING: UPDATED DATABASE
EFFECT OF DIFFERENCES BETWEEN MANUFACTURERS ON PD ACTIVITY
Vicki Warren, Iris Power Engineering
pg. 2
pg. 2
Case Study: CSMeter
Technology Averts False
Indication
IRMC 2004 Recap
pg. 4
New Field Service
Support Options
pg. 4
COURSES
September 20-23, 2004
Motor Maintenance Course
San Antonio, TX
October 26-29, 2004
Hydrogenerator
Maintenance Course
New Orleans, LA
November 16-18, 2004
Partial Discharge Course
Long Beach, CA
For a complete list of Iris’
courses, visit irispower.com
VISIT US AT THIS
YEARS IEEE PCIC
September 13 & 14
HOSPITALITY
SUITE # 1743
Partial discharges (PD) are small electrical sparks that
occur when voids exist within or on the surface of high
voltage insulation of stator windings in motors and
generators. These PD pulses can occur because of the
manufacturing and/or installation processes, thermal
deterioration, winding contamination or stator bar movement during operation. As the insulation degrades, the
number and magnitude of PD pulses will increase.
Although the magnitude of the PD pulses cannot be
directly related to the remaining life of the winding, the
doubling of PD pulse magnitudes approximately every 6
months indicates rapid deterioration is occurring. If the
rate of PD pulse activity increases rapidly, or the PD levels are high compared to other similar machines, this is
an indicator that visual inspections and/or other testing
methods are needed to confirm the insulation condition.
Furthermore, if the PD magnitudes by the same test
method from several identical windings are compared,
the windings exhibiting higher PD activity are generally
closer to failure.
The following describes the current Iris PD database and
some recent interesting findings on the effect of
manufacturer on PD levels.
COLLECTION OF DATA
PD Test Method
During normal machine operation, an instrument called the
PDA-IV or TGA is temporarily connected to the previously
installed sensors in each phase. The sensor blocks the power
frequency voltage, and conducts the high frequency voltage pulse accompanying partial discharge. To avoid any confusion with electrical
noise from power tool operation,
corona from the switchgear, RF sources, etc., the
PDA-IV or TGA separates PD from system
noise on the basis of time-of-arrival and pulse
characteristics, and measures the number,
magnitude and ac phase position of the PD pulses.
After the accumulation of all available test
data through to 2003, a total of 60,342
records, a database was carefully compiled
using only on-line tests obtained during
normal operation, and only one test result per
sensor, the latest test at Full Load and Hot stator winding temperature (FLH). Any test with
questionable results was discarded.
Number of FLH Tests by Machine Type
Motors
Hydros
Turbos
2211
4828
3953
20%
44%
36%
Number of FLH Tests by Sensor Type
Capacitors
SSCs
RFCTs
8996
1646
350
82%
15%
3%
ANALYSIS BY MANUFACTURER
13-15kV Air-cooled Machines
The database was separated based on eight winding
manufacturers using only valid full-load hot tests from
air-cooled machines with 80pF capacitors. No effort
was made to identify the manufacturing process or
insulation system. In order to make comparisons among
the results, it was necessary to divide the data further
based on the operating voltage of the specimen winding.
The focus was narrowed even further to include only
those windings that have been installed since 1990, or
windings that have been in operation fifteen years or less,
where the expectation is less partial discharge since the
winding has seen fewer operating hours.
PD results by Manufacturer
13-15kV Air-cooled Machines with 80 pF sensors
Year of Install 1990-2003
Peak Magnitude (mV)
Case Study: Delay of OEM
Recommended Rewind by
Using On-Line PD Testing
Manufacturer on PD Levels
Data Plot 1. 13-15kV Air-cooled Machines with 80pF sensors
CONTINUED ON PAGE 3
Iris Power Engineering • 1 Westside Drive • Unit 2 • Toronto • Ontario • Canada • M9C 1B2
Phone: (416) 620-5600 • Fax: (416) 620-1995 • E-mail: [email protected] • www.irispower.com
1
CASE STUDY
CASE STUDY
Delay of OEM Recommended Rewind by
Using On-Line PD Testing
CSMeter Device Technology Averts False Indication
In 1994, Montrose Unit 2, Kansas City Power and Light was taken out of
service and the OEM recommended a rewind. The utility decided against
the recommended rewind at that time and installed 80pF Bus Couplers to
facilitate PD testing. Initially, the PD was moderate. However, testing
results since 1998 showed rapid deterioration of the C phase windings,
specifically Phase C (winter operations) shows very high and unstable
PD activity. This is an indication of loose coils, poor semi-conductive
coating, causing surface discharge between stator bar and grounded core
iron. A DC Hipot test was performed and confirmed the on-line PD findings.
The CSMeter instrument is Iris’ newest product, and helps users find
broken rotor bars in squirrel cage induction motors. Several intelligent
algorithms have been embedded in CSMeter making it reliable while at the
same time removing the need for expert analysis in most cases.
The unit was taken out of service and rewound in 2003. The tests since
then show the expected low levels of PD activity. The company was able
to delay the recommended rewind for 9 years by using on-line PD
testing in conjunction with visual inspections when PD levels rose.
More details can be found in a paper by George Yeboah presented at the
2004 Iris Rotating Machine Conference (IRMC) on our website.
Phase C - Winter Operations - PD Activity since 1994
Speed-reducing gearboxes connected to a motor can induce frequencies very
close to where we expect to find broken rotor bar frequencies. CSMeter
algorithms can predict, recognize and omit such frequencies from analysis so
as to avert a false positive. This is achieved if the gearbox reduction ratios are
known and defined using CSManager software and uploaded to CSMeter
with the rest of motor nameplate data prior to initiating a test.
Such was the case at a
cement plant where a
test was performed
on a 300hp coal mill
motor rated at 575V,
295A, 885rpm, 60
Hz. The motor was
drawing a load current of 250A. The test
was done from the
MCC
using
a
1000A/1V clamp-on
current probe around one of the main phase leads to the motor after making sure that it was the unshielded portion of the cable.
CSMeter gave an instant diagnosis of "Healthy" and displayed a
frequency spectrum showing sidebands on its LCD screen after acquiring
and processing the data for approximately 3 minutes. However, none of
these sidebands indicated broken rotor bars. The graph (above) was
obtained using CSView software. Note the presence of several frequency components and absence of any reference to broken rotor bars.
CSMeter correctly omitted gearbox-induced frequencies from analysis
that would otherwise have led to a false positive.
Phase C - Data Collected Since the Rewind
PDLITEPRO & PDVIEW 3 SOFTWARE RELEASED
Iris has released that latest in data collection and viewing software for its
PDA-IV and TGA portable PD instruments. PDLitePro is the next
generation in data collection and database management tools and PDView 3 is
the latest in viewing and reporting tools from Iris.
PDLitePro includes the following enhancements plus many more:
·
·
·
·
·
·
Communication with USB port, RS232 and Ethernet (TCP/IP and DCHP)
Multiple test configurations can be stored before heading out into the field
Auto-generate sensor configurations
Change/modify operating parameters after a test is taken
Supervisory functions which allow a "lock down" of the application
Compatible with Windows 98 and above
™
PDView includes these new features just to name a few:
· New plots to enhance the ability to determine if PD is occurring in
the slot or endwinding for Bus coupler and PDA installations
· More standard report templates for PD analysis comparable to
AdvanceView
· Improved data filtering capabilities for trend plots
· For the expert user, there are new optional analysis tools in the
advanced summary numbers (such as mode, mean, skew, etc.)
· Compatible with Windows 98 and above
™
For more information on obtaining PDLitePro & PDView and for
upgrading your PDA-IV or TGA instrument, please contact your sales
representative.
Iris Power Engineering • 1 Westside Drive • Unit 2 • Toronto • Ontario • Canada • M9C 1B2
Phone: (416) 620-5600 • Fax: (416) 620-1995 • E-mail: [email protected] • www.irispower.com
2
PARTIAL DISCHARGE TESTING
continued from front cover
In data plot 1, it is apparent that manufacturers A, D, E, H and J have
significantly lower PD than the other manufacturers. The percentile refers to
the percentage of tests that have a PD magnitude below that indicated. For
example, 95% of readings for Manufacturer B have a peak magnitude (Qm)
below 1800mV. The reason for the higher partial discharge levels from some
manufacturers may be inherent to the insulating materials and not indicative
of a problem with the manufacturing or installation process. However, for
manufacturers B, G, I and Misc, at least 10% of their machines have Qm
levels above the expected levels for similar windings, and therefore, indicate
a problem exists within the winding due to either the design, manufacturing,
installation, or operating history of the machine.
As illustrated in Data Plot 1 on page 1, it is evident that with the exception of
Manufacturer B, most of the results are similar among the manufacturers.
Note that at least 75% of the machines for all the manufacturers, and up to
90-95% for some, have normal PD levels for this operating voltage, and therefore, are well-designed, well-made, and well-installed. In general, to have a
problem in a winding is the exception to the rule, perhaps merely 5-10% of
the machines manufactured.
CONCLUSION
Though it is always recommended that you trend the results for one machine over
time and thus monitor the rate of degradation of the stator winding, it is also
possible to compare results from similar machines provided the following
conditions remain the same including test instrument, type of sensors, operating
voltage of the machines and operating gas environment of the machines.
The time of winding failure is normally the result of a deteriorated winding
being subjected to an extreme stress such as a lightning strike, out-of-phase
synchronization, excessive starts, or system imbalance. As these are
unpredictable, it is impossible to forecast when a failure will occur. However,
by monitoring the PD characteristics of a stator winding, it is often possible to
determine which machines are more susceptible to failure and therefore which
require maintenance.
8-12kV Air-cooled Machines
The results of a similar analysis for machines operating between 8-12kV are
displayed in Data Plot 2. As with the 13-15kV machines, there is significant difference in the PD levels between the manufacturers. The 90th percentile from
manufacturers A and B is almost twice the PD levels of the other manufacturers,
and three times that of C. When comparing the results of only the newer machines,
there is significant difference throughout the spectrum indicative of wide variances
in manufacturing materials and processes.
It is apparent that two manufacturers, A and B, have higher PD than the
others to indicate a problem within the winding due to either the design,
manufacturing, installation, or operating history of the machine. For all
manufacturers at least 50-75% of the machines for the manufacturers, and up
to 90-95% for some, have normal PD levels for this operating voltage, and
therefore, are well-designed, well-made, and well-installed.
Data Plot 2. 8-12kV Air-cooled Machines with 80pF sensors
4-7kV Air-cooled Machines
The results of a similar analysis for machines operating between 4-7kV are
displayed in Data Plot 3. Similar to the 13-15kV machines, there is
significant difference in the PD levels between the manufacturers. The 90th
percentile from manufacturer D and J is almost twice the PD levels of those
of manufacturer A, and three times E and H. When comparing the results of
only the newer machines, there is significant difference throughout the
spectrum indicative of wide variances in manufacturing materials and
processes.
It is apparent that two manufacturers, D and J, have higher PD than the
others. Manufacturer A also has noticeably higher PD, but lower than D and
J. With the exception of manufacturer J, at least 75% of the machines for the
manufacturers, and up to 90-95% for some, have normal PD levels for this
operating voltage, and therefore, are well-designed, well-made, and
well-installed.
Data Plot 3. 4-7kV Air-cooled Machines with 80pF sensors
Plan to attend next years IRMC
SCOTTSDALE, ARIZONA
For more information, please contact KIM ZARB at
[email protected] or 416-620-5600 extension 240.
Iris Power Engineering • 1 Westside Drive • Unit 2 • Toronto • Ontario • Canada • M9C 1B2
Phone: (416) 620-5600 • Fax: (416) 620-1995 • E-mail: [email protected] • www.irispower.com
3
UPCOMING EVENTS
ce
, Fran
, Paris
E
R
IG
pt 3: C
, CA
9 - Se
2
g
ncisco
u
a
r
A
F
n
Sa
PCIC,
IEEE
:
6
1
3
lis, IN
Sept 1
ianapo
d
In
,
4
EI 200
-22: IS
9
1
t
p
Se
Septe
mber
20-23
: Moto
San A r Maintena
ntonio
nce C
, TX
ourse
Octob
,
er 2629: Hy
Main
tenan drogener
New O ce Cours ator
e,
rleans
Novem
, LA
ber 16
-18: P
art
San A ial Dischar
ge Co
ntonio
, TX
urse,
FIELD SER VICE SUPPORT T O CLIENTS AND AGENTS
The field service group now offers two types of support to agents and
customers, this support is available from 10:00 am to 3:00 pm (Eastern
North America). Support outside of these hours is not guaranteed and is
subject to availability of staff. It may be possible to schedule a specific time
outside of these hours as required.
PRODUCT SUPPORT: James Thompson provides product support
relating to hardware and software issues. He can be reached at
[email protected] or at 416-620-5600, extension 229.
TECHNICAL SUPPORT: Field Service Specialists are available to
provide technical support and advise on the design and/or installation of
Iris products. As this role rotates and our specialists travel throughout the
world, it is essential that all correspondence goes though
[email protected] to enable us to provide continuous support. You
may also call 416-620-5600, extension 272 with any questions.
If you are unsure as to what support is required please contact Product
Support and your call will be forwarded to the appropriate department.
Should either of the above not be available or if you wish to follow up
with any support, please contact Andy Brown, GM Field Service at
[email protected], or 416-620-5600 extension 295.
2 0 0 4 I R I S R O TAT I N G M A C H I N E C O N F E R E N C E R E C A P
Once again we have had another successful IRMC. This year we returned to that
wonderful city New Orleans and held the conference in the JW Marriott (for those of
you who were at the last conference in New Orleans - this is the old Le Meridian).
True to form the weather was hot and humid and Bourbon Street was a party every
night.
We had people from all corners of the world including India, France, Japan, Malaysia,
South Africa, Australia, Canada and USA.
The program was divided into two training
sessions and two technical sessions, breakout
sessions and a walk-in clinic. The wide array of
technical papers provided something for each and every attendee. As is customary, the training sessions were full;
customers were introduced to the newly released CSMeter, PDLitePro and PDView3 in the users group meeting;
PDView3 and PDLitePro were reviewed in more depth in a separate session; and at the walk-in clinic machines were
given a clean bill of health. These sessions are a great opportunity to learn all about condition-based maintenance and
Partial Discharge theory for both new and seasoned users of the technology.
will
rs IRMC
Next yea ne 2005 in
u
J
be held
ale, AZ
Scottsd
We would like to thank all the authors for their time and effort in presenting very informative technical papers; the
session chairs for maintaining the flow of the program; and most of all thank you to all those who attended.
Next year's conference will be held in Scottsdale, Arizona - watch our website for specific dates. Anyone interested in
presenting a paper, please email [email protected]
Iris Power Engineering • 1 Westside Drive • Unit 2 • Toronto • Ontario • Canada • M9C 1B2
Phone: (416) 620-5600 • Fax: (416) 620-1995 • E-mail: [email protected] • www.irispower.com
4
Fly UP