0% found this document useful (0 votes)
217 views4 pages

Insulator Testing of AC Generator Winding

IEEE journal publications
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
217 views4 pages

Insulator Testing of AC Generator Winding

IEEE journal publications
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

2017 Electrical Insulation Conference (EIC), Baltimore, MD, USA, 11 - 14 June 2017

Insulation Testing of AC Generator Windings


A Review of Tests Typically Performed by Hydro-Plant Maintenance Staff

Scott Cotner, P.E.


Hydroelectric Design Center
US Army Corps of Engineers,
Portland, OR USA

Abstract— The useful life of an electrical generator is directly remove the turbine runner for overhaul or replacement may
related to the condition of the winding insulation. Non-invasive require that the rotor be removed from the generator. Having
methods to assess the condition of insulation systems are typically the service-aged generator fail a few years after the turbine
limited to physical inspections, drawing inferences from a review runner replacement would result in two lengthy outages when
of operating conditions and maintenance records, and from the the work could have been performed in a single outage.
results of on-line and off-line electrical testing. This paper Performing condition assessment testing of the generator
identifies when to perform off-line electrical testing, notes testing before starting the turbine runner replacement work allows
standards and fundamentals, provides a description of the appropriate steps to be taken to ensure the generator and the
commonly performed tests and pass/fail criteria, supplies some
runner both have long, reliable service lives at the conclusion
framework to aid project staff in deciding over-potential test
voltages, and articulates the implicit trade-off in over-potential
of the capital refurbishment work
testing.
C. Winding Insulation Failure
Keywords—insulation testing, maintenance tests, insulation A dielectric failure of a generator stator or field winding
resistance, pole drop, DC ramp, test voltages forces the generator out of service until the damage is
repaired. The failure provides several opportunities to test the
I. INTRODUCTION insulation system. Diagnostic testing is performed to
This paper presents an overview of maintenance testing of determine the extent and location of the damage, which can
hydro-electric generator insulation, including when testing guide the development of repair options. Testing is performed
should be performed, testing standards and fundamentals, on the “healthy” portions of the winding after the failed
descriptions of commonly performed tests and results portion has been isolated to confirm that no other weak
interpretation, and factors to consider for testing at voltages insulation is present. Testing is also performed on the repaired
above winding ratings. The electrical testing that is typically winding to provide a measure of assurance of continued
performed in the field tends to be limited to ground insulation reliable operation.
testing for the stator winding and ground and turn insulation
for the field winding. References 8 and 15 provide further III. TESTING STANDARDS AND FUNDAMENTALS
information on testing theory as well as other tests that could
be performed during maintenance outages. A. Insulation Testing Standards
IEEE standards cover electrical testing of electrical
II. WHEN TO PERFORM INSULATION TESTING insulation systems. The standards listed in the references
comprise those most used or referenced in field testing of ac
A. Routine Maintenance generator insulation systems [1-8].
Electrical insulation testing should be included as a part of
the normal, time-based, maintenance performed on electrical B. General Testing Methods
generators as the test results provide a snapshot of the A clear understanding of the insulation system components
insulation system condition. Trending of the results of periodic intended to be tested is required when planning the testing
tests can provide insight into the effects of changing operating connection configurations, particularly in cases where the
conditions (e.g., higher winding temperatures) or even provide winding has been sectionalized to isolate an insulation failure.
warning that more intensive maintenance activities may be Care must be taken to ensure only the intended components
required for the generator to continue operation (e.g., partial will be energized by the test.
generator disassembly and cleaning of a field winding).
In all types of electrical insulation testing, it is preferred
that the test voltage be applied to the winding and increased
B. Condition Assessment from zero to the test voltage smoothly. The winding under test
When planning significant capital investments, a condition is held at that voltage for the specified time period (usually one
assessment effort is undertaken to better characterize the minute), and then the test voltage is smoothly brought back
condition of the major plant components to ensure that down to zero [1]. Notable exceptions to this method of
investments are made wisely. For example, the process to

978-1-5090-3967-8/17/$31.00 ©2017 IEEE 447


applying the test voltage include the Polarization Index test and The test results should be compared to either the
the stepped and ramp dc over-potential tests [4]. recommended IR minimum of “winding kV +1”, corrected to
40°C, minimum PI is 1.5 or 2 (depending on winding
Adjustable sphere gaps or some other overvoltage insulation type) or manufacturer’s recommended minimums
protection should be used to prevent test set failures from and the previous test results for the same equipment
applying much higher than intended voltages to the winding referenced to the same temperature [2]. If the latter are not
under test [3, 5]. available, the current or previous results from similar
generators in a similar duty may be used for comparison. Low
C. Grounding After Testing or decreasing PI results can indicate a dirty or wet winding
Winding insulation systems have significant capacitance (contaminated insulation surface).
and can retain considerable energy even after the test voltage
has been reduced to zero and the winding under test is Trending the results of IR/PI tests over many test intervals
grounded. A typical recommendation to safely drain the stored can provide information about long term degradation of
energy is to ground the winding for 4 times as long as the test winding dielectric strength or condition and allow planning of
voltage was applied, or at least 60 minutes [3]. Grounding is corrective intervention. Note that field winding PI results are
typically performed with a resistive ground tool applied with often low due to lower capacitance of the insulation materials
insulated gloves as the windings must be considered involved [1]. As passing the IR/PI test suggests the winding is
dangerous until completely discharged. [4] not dirty or wet, the test is often used as a prerequisite for
performing higher voltage insulation tests [4].
D. Safety/Barricading 2) Pole Voltage Drop Test
Electrical testing should be performed only by experienced A pole voltage drop (or impedance) test is typically
and qualified personnel as hazardous energy is involved. The performed on field windings only. An ac voltage of 120V or
area of the testing should be electrically isolated and 240V is applied across the field winding at the field leads or
barricaded to prevent ingress by unauthorized persons. Project the collector rings. The voltage drop across each field coil or
or Plant specific safety rules regarding signage, access control, pole is measured and recorded, as is the total applied voltage
and other aspects of high voltage testing safety should be at the start and end of the test. The pole voltage readings are
followed. then averaged, and any significant deviation (typically 5%) of
a single pole reading to the average is investigated.
Temporarily shorting one turn of a field coil turn to prove that
IV. COMMONLY PERFORMED INSULATION TESTS AND
the test will detect a coil with shorted turns is recommended
THEIR INTERPRETATION [3, 5, 6].
A. Tests Performed Below Rated Winding Voltage A shorted field coil will reduce the readings of the adjacent
The tests that are typically performed at below rated healthy coils significantly. Typical readings may be 30-40%
winding voltage include the Insulation Resistance and of the average voltage on the pole with the shorted coil and
Polarization Index test, the pole drop test, and the ac 60-70% of the average voltage on the adjacent (healthy) coils.
impedance test. The choice of test voltage for these tests does Confirming the influence of the pole with the shorted turn on
not change with the age of the winding to be tested as none of the adjacent poles can be performed by electrically bypassing
the test voltages should pose any danger to serviceable the suspect pole. [3]
insulation.
B. Tests Performed Above Rated Winding Voltage
1) Insulation Resistance/Polarization Index
The tests that are typically performed at above rated
The most commonly applied test for both stator and field
winding voltage include the surge test and dc and ac over-
winding insulation is the Insulation Resistance and
potential tests. The choice of test voltage for these tests must
Polarization Index test (IR/PI). The test involves applying a
consider the age of the winding and the goals of the test, as the
steady dc voltage to the winding insulation and measuring the
test voltages can damage serviceable insulation.
resulting current; the ratio of the two is the winding insulation
resistance. As described in [1] and [8], the winding resistance 1) Surge Testing
is calculated after 1 minute and after 10 minutes of the Turn insulation on multi-turn stator or field windings can
application of the test voltage. The ratio of the 10 minute to be tested singly or in small groups of coils using a surge test.
the 1 minute insulation resistance (IR) is the polarization The surge test applies a voltage pulse to the coil or coils, and
index (PI). While the IR is dependent on temperature and records the resulting oscillatory current wave. For surge tests
requires correction to compare one test to another, the PI test results, the results for one coil are compared to those of
results can be compared across multiple tests. identical “healthy” coils. By comparing the wave shape to that
The IEEE recommended test voltages for the IR/PI tests of many other identical coils or groups of coils, deviations
are 5kV or 10kV for a 13.8kV stator winding and 250 or 500V from the norm can be detected and coils with shorted turn
dc for a field winding [5]. It is preferred that the same test insulation identified.
voltage be used for periodic tests to allow comparisons among More involved surge test methods are available for testing
the results. coils in a complete stator winding by coupling surges to the

448
coils using a full size surge inducing coil, but those tests are b) AC Over-potential Tests
infrequently performed in the field [3]. The ac over-potential test is similar to the dc over-potential
2) Over-potential or “Hipot” Tests test except that the applied test voltage is 60Hz ac voltage.
Over-potential tests, or “hipot” tests, involve the The windings are prepared the same way, and the test voltage
application of higher than rated voltage between the winding is raised smoothly from zero volts to the test voltage, held
conductors and ground with the intent of exposing insulation there for 1 minute, then smoothly ramped back down to zero
flaws that could cause a near term in-service failure of the volts. The winding passes the test if the test voltage is
insulation. With the exception of the stepped/graded and ramp sustained without a disruptive failure of the insulation or a
dc test, all over-potential tests are of the proof type (“go/no- tripping of the test set overcurrent protection. Little diagnostic
go”) and offer little diagnostic or condition assessment information can be gleaned from the test [3, 5].
information beyond the fact that the winding insulation could c) AC or DC Over-potential Testing?
withstand the test voltage on the day of the test [4]. The test is
performed to provide some assurance that the winding is There are several factors involved in the choice between
unlikely to suffer an insulation failure in operation in the near performing ac or dc over-potential tests. First is size and cost
term assuming normal slow-acting insulation system of the equipment. The dc test equipment is much smaller, less
deterioration causes. An invasive repair and retest is required expensive and more portable than ac test equipment. Second,
before the generator can return to service if the winding does the way the applied test voltage is distributed across the
not withstand the test voltage. winding’s ground insulation is different. In the dc test, the
voltage is distributed across the insulation in proportion to the
a) DC Over-potential Tests resistivity of the components. In the ac test, the voltage is
A conventional dc over-potential test is conducted in the distributed across the insulation in proportion to the
same manner as the IR test. The windings are prepared the capacitance of the insulation components. The result is a
same way, and the test voltage is raised smoothly from zero different voltage distribution (note that neither are equivalent
volts to the test voltage, held steady for 1 minute, then to the in-service voltage distribution). In modern epoxy mica
smoothly ramped back down to zero volts. The ramp rate of the insulation systems, the dc test may not detect flaws or
test voltage is limited by the current rating of the test weaknesses that the ac test could detect [4]. Another
equipment, typically between 5 and 25 mA. The winding consideration with ac over-potential testing is the application
passes the test if the test voltage is sustained for one minute of an ac voltage equal to or in excess of the winding’s rated
without a disruptive failure of the insulation or a tripping of the voltage may induce partial discharges that will age the
test set overcurrent protection. Little diagnostic information winding slightly [8].
can be gleaned from the test. Over-potential test voltages for dc testing are determined
by choosing the desired ac test voltage and multiplying by 1.7
Variations on the conventional dc over-potential test [4, 5, 7]. There is some suggestion that the factor may not be
include the fixed and variable interval stepped test and the dc appropriate for epoxy-mica insulation because the factor was
ramp test. The most recently developed test is the dc ramp test. developed based from test results using asphalt-mica insulation
In this test, an automatic tester raises the test voltage at a systems [8], or because it fails to account for the effects of dc
steady rate (usually) 1kV/minute to render the capacitive voltage on stress grading systems [12].
current constant over the period of the test [6, 9, 10].
For the field winding, ac tests are preferred [7]. The field
The purpose of all three variations on the conventional dc winding is subject to contamination from the brushes and
over-potential test is twofold. First, by plotting the applied bearing oil mist and the winding is normally excited with a dc
voltage and measured current, a characteristic curve can be voltage. The contamination levels increase the likelihood of
developed and compared to future (or past) test results. The tracking developing during an extended dc over-potential test,
second purpose is to provide real time feedback to the test and the normal dc voltage on the field will make flashovers to
technician or engineer of a sudden change in current, thus ground more likely along those possible tracking paths. The
possibly allowing the test to be stopped before an insulation lower recommended test voltage reflects the less severe
rupture occurs. dielectric duty of the rotor winding insulation and, therefore,
the lower capabilities of the rotor winding insulation.
There are two components to interpret the stepped/graded
and ramp dc test results. The first is whether or not the winding
withstood the intended voltage without failure or insulation V. SELECTION OF OVER-POTENTIAL TEST VOLTAGE
rupture. This is a “go/no-go” portion of the test results. If the Offline electrical insulation testing cannot induce dielectric
winding withstood the applied voltage, then the proof test stresses in the insulation system in the same manner as is
portion of the test goal was met. found during normal generator operation. For a stator winding,
The stepped/graded and ramp dc test results also provide the application of any test voltage to a winding results in the
current versus time/voltage curves that are useful in inferring same voltage to ground at all points along the winding, but in
the condition of the winding. The curves are analyzed for slope operation the voltage to ground decreases as the measurement
and linearity, with particular attention paid to changes in either. point moves away from the terminal (i.e., toward the neutral
There are several references in the literature that provide of a Wye connected stator winding). One result of this
interpretive recommendations based on curve shapes and post- difference is that a winding with a dielectric weak point in the
testing coil dissections of healthy and failed coils [4, 9 13, 14]. insulation nearer the neutral connection where the normal

449
operating voltage is low may fail an over-potential test, yet the VII. CONCLUSION
winding may not have failed in service. Routine electrical testing provides information on the
As these tests are performed at potentials higher than the evolving condition of generator winding insulation and some
rated voltage of the winding under test, some consideration of assurance that the insulation system will operate reliably into
the age of the winding and the desired information to be the near future. Electrical testing also allows insulation
gathered from the test should be considered. weaknesses to be identified under controlled conditions where
the impacts of an in-test failure are reduced considerably when
For field windings, recommended over-potential test compared to in-service failures. On the other hand, testing can
voltages are limited to ten times the rated field voltage, but no result in an insulation failure that may not have occurred in
less than 1500Vac [7]. New field windings may be tested at actual operation due to differences in voltage distribution
voltages between 5000V and 7000Vac, but service aged field across the winding in the two different conditions.
windings are typically tested at voltages between 1500 and
2500Vac. It is not common to perform over-voltage testing on Whether to perform and what test voltage to be applied in
service aged field windings [11]. over-potential testing needs to be carefully evaluated to
balance industry standard recommended values against the
The range of recommended test voltages for ac over- goals of the testing to minimize unnecessary insulation
potential tests for stator windings is 1.5 to 1.25, or in some failures while still producing meaningful test results.
cases as low as 1.1, times the voltage rating of the winding
[3,11]. Relatively new windings would be tested at the high
end of the range, service aged windings near the middle and REFERENCES
windings near the end of their service lives or with known
winding insulation weaknesses would be tested at voltages [1] IEEE Standard 43 – IEEE Recommended Practice for Testing
Insulation Resistance of Rotating Machinery.
nearer the lower end of the range. The choice of over-potential [2] IEEE Standard 56 – IEEE Guide for Insulation Maintenance of
test voltage is one of risk tolerance. Large Alternating-Current Machinery (10 000kVA and Larger).
[3] IEEE Standard 62.2 – IEEE Guide for Diagnostic Field Testing of
Electric Power Apparatus – Electrical Machinery.
VI. IMPLICIT RISK-BENEFIT TRADE-OFF IN OVER-
[4] IEEE Standard 95 – IEEE Recommended Practice for Insulation
POTENTIAL TESTING Testing of AC Electric Machinery (2300V and Above) With High
All over-potential insulation testing must address this trade- Direct Voltage.
[5] IEEE Standard 115 – IEEE Guide: Test Procedures for
off: the risk of a possible insulation failure during testing with Synchronous Machines.
low follow-on current and minimal winding damage but at a [6] IEEE Standard 492 – IEEE Guide for the Operation and
test voltage higher than the winding would ever see in service Maintenance of Hydro-Generators.
vs. a possible in-service insulation failure with possibly large [7] IEEE Standard C50.12 – IEEE Standard for Salient-Pole 50Hz and
follow-on currents that could cause substantial winding 60Hz Synchronous Generators and Generator/Motors for
Hydraulic Turbine Applications Rated 5MVA and Above.
damage. The equipment owner’s evaluation of this risk-based [8] Greg C. Stone, Edward A. Boulter, Ian Culbert, Hussein Dhirani,
decision directly affects the magnitude of the test voltage Electrical Insulation for Rotating Machines, IEEE/John Wiley
chosen, and possibly even what testing is recommended [3]. Interscience, 2004.
[9] Rux, L. M., “The Physical Phenomena Associated with Stator
Establishing the goal of the electrical testing informs the Winding Insulation Condition as Detected by the Ramped High-
decisions regarding which tests to perform and to what test Voltage Method”, Ph.D. thesis, Mississippi State University, May
voltages the winding insulation will be subjected. For 2004.
example, a higher test voltage may be chosen if the test is [10] Rux, L.M., “High Voltage DC Tests for Evaluating Stator Winding
intended to provide some assurance that the winding Insulation: Uniform Step, Graded Step, and Ramped Test
Methods”, presented at IEEE Conference on Electrical Insulation
insulation will perform reliably for 3-5 or more years. A lower and Dielectric Phenomena, Minneapolis, MN, 1997.
test voltage level may be chosen if the winding needs to only [11] Greg C. Stone, Relu Ilie, “Turbine Generator Rotor and Stator
perform reliably for a short period of time before a stator Winding Hipot Testing”, IEEE Electrical Insulation Magazine,
rewind is started. Vol. 28, No. 21, IEEE Dielectrics and Electrical Insulation
Society.
There are some special cases where returning a generator [12] Emery, F.T., “Considerations of AC and DC Testing of Large
to service with poor test results may be worthwhile. One case Electrical Stator Windings and Components”, IEEEXplore.
could be when the period of operation between the test and the [13] Bertenshaw, D. R. and Rux, L. M., “Assessing the Condition of
Rotating Machine Stator Winding Insulation Using the Ramped
date of a major repair or refurbishment (like a stator winding Direct High-Voltage Method”, presented at INSUCON 2002,
replacement) is short and the value of the generation is high. Berlin, Germany.
Another case could be that the test results indicate a [14] Eastment, E. and Milano, B., “30 years of Ramp DC Testing
deteriorated condition which normal operation could improve. Hydrogenerator Stator Windings Test Results and Interpretations”,
For example, a low IR and PI reading on a field winding could presented at the 2008 EPRI Generator Predictive and Preventative
Maintenance Workshop.
indicate that the surface of the insulation on the field winding [15] Stone, G. C., “Recent Important Changes in IEEE Motor and
is damp. Normal operation of the generator raises the Generator Winding Insulation Diagnostic Testing Standards”,
temperature of the field winding, which would drive off the IEEE Transactions on Industry Applications, Vol. 41,
surface moisture. The insulation must be capable of January/February 2005.
withstanding normal operating voltages for this to be a viable
option.

450

You might also like