High-potential (hipot) withstand testing, which indicates whether the stator winding insulation in a generator is fit for service, is a standard factory acceptance test for a new machine. The same test can be used on an existing generator to assess the machine’s condition. EPRI surveyed owners of generating facilities about hipot testing on existing machines. This article answers questions raised during the survey, including when to test, the type of test to perform, and the proper test voltage.
By Jan Stein, Greg C. Stone, and Bal K. Gupta
High-potential (hipot) withstand testing is a tool hydro plant owners can use to ensure the high-voltage insulation on the generator stator winding is fit for service. During a hipot test, performed with the generator off line, a voltage 100 to 200 percent higher than the normal operating voltage is applied between the stator winding copper conductors and the core. If the winding insulation does not fail (i.e., puncture) during this test, this indicates the stator winding likely will not fail when it is returned to service or for the next few years. A sufficiently high test voltage is selected to ensure the winding will remain serviceable for about three to five years.
Although all generators must pass a hipot test before being commercially accepted from the manufacturer, many plant owners do not use hipot testing once the winding is in service. In particular, owners are concerned that performing this test may lead to deterioration of the insulation, even if it does not fail. In addition, project owners do not know the proper test voltage to use that will allow them to find important insulation problems but not cause a nuisance breakdown. Finally, there are several types of hipot tests (see sidebar on page 38), and engineers are uncertain about how to select which test to use.
To address these issues, EPRI conducted two surveys of plant maintenance personnel from many utilities – one in 2000 and one in 2007 – to determine their practices regarding hipot testing. EPRI also conducted an extensive literature review to address some of the technical issues surrounding hipot testing of generators. This article describes the results of this research and addresses some of the concerns that arose during the survey.
EPRI’s hipot testing survey
In 2000, EPRI mailed a questionnaire on hipot testing to about 60 industry experts, including utility engineers, original equipment manufacturers (OEM), and service organizations. EPRI received 45 responses from 32 utilities in North America; five utilities in Asia, Africa, and Europe; two utility service providers; five OEMs; and one retired engineer from a major European utility. The results of that survey were published in an EPRI report.1
Because of a perceived change in utility attitudes with regard to applying maintenance hipot tests, EPRI repeated the survey in 2007 to determine the changes in attitudes that have actually occurred. This survey was sent to all EPRI member utilities, and 29 utility maintenance engineers responded. The details of the survey also were published in an EPRI report.2
With regard to new windings, the results of both surveys were quite clear. All the respondents required a hipot test before commissioning of the unit. Almost all respondents to both surveys indicated that they prefer to use the alternating current (AC) hipot test, using specified test levels.3, 4 Table 1 on page 36 shows the comparative advantages and disadvantages of AC and DC hipot tests.
However, when it comes to performing hipot tests for maintenance purposes, there was considerably more diversity:
– Three-quarters of respondents use either an AC or direct current (DC) hipot test for maintenance purposes. Thus, only 25 percent do not perform hipot testing once the generator has been operating (a much smaller percentage than occurred during the 2000 survey).
– About 40 percent of all respondents used AC hipot tests for maintenance purposes. A quarter of respondents use AC tests exclusively. The rest use a mixture of AC and DC tests. The reasons respondents gave for preferring AC tests over DC tests are: stress distribution across the insulation is similar to that experienced when the unit is in service; the test provides better efficacy in detecting insulation weakness; they were recommended by an OEM; and use of the same power supply for diagnostic measurements such as partial discharge (PD) and dissipation factor (DF).
With regard to voltage levels for AC maintenance hipot tests, most survey respondents use 60 to 80 percent of 2 E plus 1 kilovolt (kV) (where E is the rated root mean square [rms] phase-to-phase voltage of the stator) for maintenance tests, depending on age, condition, etc. However, two respondents used voltage as low as 1.0 to 1.1 E kV for machines that are more than a few decades old.
– As stated before, three-quarters of respondents use DC hipot tests at least to some degree. Results showed about 40 percent use the conventional test, 45 percent the DC step (leakage) test, and about 15 percent the ramp test. (It should be pointed out that the ramp test is not only a hipot test but is also a very sensitive diagnostic tool.) The main reasons survey respondents gave for preferring DC hipot tests are: small and inexpensive DC supplies; sometimes there is a warning that enables the test to be aborted before a weak winding is punctured; and the DC test may produce less consequential damage to the insulation than the AC test (according to 17 percent of respondents). However, only 7 percent of respondents believe a DC test is better than an AC test at detecting an insulation weakness.
Table 1: Comparison of AC and DC Hipot Tests1
For DC hipot tests, about half of respondents prefer a maintenance test level of 60 to 80 percent of 1.7 (2E + 1) kV. This works out to about 2 E, applied phase to ground. However, many other values – such as 1.13 E, 1.25 E, and 1.7 E – also were used.
– Use of the very low frequency (VLF) or 0.1 Hz hipot tests is rare. Just 3 percent of the respondents indicated they use these types of hipot test, but only if the generator OEM prefers it.
Important hipot issues
The objective of the survey was to answer some important questions on the use of hipot tests. Here, several queries are addressed, using findings from the survey of industry experts and existing literature.
When should you apply hipot tests?
For new coils and machines, hipot tests are used universally as acceptance and quality assurance tests. Preferably, other diagnostic tests – such as insulation resistance (IR), polarization index (PI), DF, and/or PD – should precede hipot tests in the detection of serious flaws for a new machine.5,6,7 The AC hipot test should be required for the acceptance hipot because it tends to be more sensitive to manufacturing flaws at test levels set by the standards.3,4
Maintenance hipot tests can assure hydro project owners that the stator winding will not fail in-service during the next several years as a result of insulation aging. In essence, the decision to perform a maintenance hipot test requires a management assessment that weighs the overall cost of an in-service failure vs. either the assurance provided by a successful hipot test or the cost of a hipot test failure. Failure of the winding during a hipot test generally indicates that it likely would have failed in service (perhaps during a power system transient). Thus, use of a maintenance hipot test is prudent if the cost of a loss of production as a result of an outage could be significant. Factors to be considered include the nature of the application of the machine (i.e., how much would a long outage cost in terms of replacement power), redundancy built into the plant for continued operation with a failed machine, availability of a spare machine for quick replacement, and insurance implications (some insurance companies will not allow a “business interruption” claim in the event of a hipot test failure).
Do hipot tests damage windings?
This question may be raised by plant managers who must approve the tests. The answer is no – maintenance hipot tests do not degrade a good winding. Even theoretically, the insulation in a good machine should not suffer any detectable degradation during a hipot test.5,7,8 All coils and bars used in modern hydro generators have the capability to pass a voltage endurance test.9 Typically, coils for a 13.8-kV winding should survive for 250 hours at 35 kV (or 400 hours at 30 kV) at about 120 degrees Celsius (C). If a 400-hour test at 30 kV represents 25 years of life in service, then a 1 minute 29 kV rms AC overpotential test ages the insulation by just nine hours – which is insignificant.
However, there is a small risk that a marginal winding that would have operated for some time in service may be punctured and prematurely fail by using the hipot test. For example, a winding with poor coils near the neutral end may operate in service for many years but may fail a hipot test. However, machines that fail a hipot test typically have poor insulation. Chances are that these machines would have failed in service, especially in the event of an overvoltage from surges or a power system fault, because even the neutral end coils may see a temporary high voltage for a short time.
The hipot failure will require immediate repairs or replacement and delay return to service. This risk can be minimized (but not eliminated) by using the DC ramp or step versions of the hipot test, together with other diagnostic tests (IR, PI, PD, and/or DF), which may detect insulation problems without producing a puncture.7 On the plus side, a winding that fails under test during a scheduled outage is comparatively easy and inexpensive to repair as compared to most in-service failures. This is especially true for multi-turn windings. This is a strong incentive for performing diagnostic tests during scheduled maintenance outages.
How do you choose a test?
IEEE standard 95 and other sources discuss the relative merits of AC and DC hipot tests and their main differences.1,5,7,8,10,11,12 From these differences and the survey results, the question on the use of AC or DC hipot tests can be answered.
The controversy about the relative efficacy of AC and DC hipot tests for detecting insulation weaknesses has continued for years.1,2 However, the electric stress distribution across the thickness of the groundwall insulation in operating machines is the same in an AC hipot test and different in the case of a DC hipot test. As a result, AC tests appear to be more effective in detecting defects in slot sections, while DC hipot tests appear to be more effective in detecting defects in end windings.1,5,7 Some defects in end windings that may fail in DC hipot tests may never fail under operating AC stress.1,2M However, experience shows failures in the end turns under any test to be extremely rare. Further, failures in this area are more easily repaired. The survey shows that some engineers and managers fear that more damage will occur to the stator core in a failure during an AC hipot test than during a DC hipot test. However, experts using AC hipot tests report that they have never experienced core damage from a test failure.8 The semi-conductive slot paint and coil or bar semi-conductive coatings serve to protect the stator iron from damage.
It is best to first perform other diagnostic tests (IR, PI, DF, and PD) before performing the hipot test. If bad results are obtained from these tests, the hydro project owner can then decide not do the hipot test, make repairs, or use a lower hipot voltage for the test.
Although some engineers consider the AC hipot test to be technically superior, performing a DC hipot test is better than no test at all. One advantage of the DC controlled-voltage test is that abnormalities in the current vs. voltage plot may provide a warning of an impending failure. The use of current vs. voltage plots for detecting voids and delamination is somewhat subjective,13,14,15 but it is an excellent diagnostic tool for detecting abnormal leakage currents in windings and has found problems in new and old windings not detected via the standard AC or DC hipot test.
If the past history of a machine indicates problems in the end windings, both AC and DC hipot tests may be conducted. For large machines, where an adequate AC supply may not be available, a DC hipot is the only currently used alternative. However, the VLF test technology is rapidly evolving, so it may be an alternative in the future. For water-cooled machines, AC hipot tests are preferred because DC hipot tests should not be applied without thorough drying.
What is the proper hipot test level?
For maintenance tests, the most common AC hipot level is about 1.5 E, 2 E is most common for a DC test, and the VLF test would be performed at about 1.63 times the AC hipot level (2.4 E). If the insulation is known to be in poor condition, as indicated by other tests (PD, IR, PI, and/or DF), the test level is usually lower and determined based on risk of failure.
Mr. Stein may be reached at EPRI, 3412 Hillview Avenue, P.O. Box 10412, Palo Alto, CA 94303; (1) 650-855-2390; E-mail: email@example.com. Dr. Stone may be reached at Iris Power LP, 3110 American Drive, Mississauga, Ontario L4V 1T2 Canada; (1) 416-620-5600; E-mail: firstname.lastname@example.org. Mr. Gupta may be reached at AOL Technologies Inc., 45 Rossburn Drive, Etobicoke, Ontario M9C 2P9 Canada; (1) 416-621-6035; E-mail: email@example.com.
- Guide for Rotating Electrical Machine Hipot Testing, EPRI Report 100666, EPRI, Palo Alto, Calif., 2000.
- Guide for Rotating Electrical Machine Hipot Testing: 2007 Update, EPRI Report 1014908, EPRI, Palo Alto, Calif., 2008.
- Rotating Electrical Machines – Part 1: Rating and Performance, IEC 60034-1, International Electrotechnical Commission, Geneva, Switzerland, 2002.
- Motors and Generators, NEMA MG 1-2006, National Electrical Manufacturers Association, Rosslyn, Va., 2006.
- Draper, R.E., and R.H. Rehder, Overpotential Testing of Insulation in Hydro Generators, IEEE Publication 97TP119-0, Institute of Electrical and Electronics Engineers, New York, N.Y., 1997, pages 10-13.
- Emery, F.T., Electrical Testing of High Voltage Stator Coils from Westinghouse Perspective, IEEE Publication 97TP119-0, Institute of Electrical and Electronics Engineers, New York, N.Y., 1997, pages 7-9.
- Stone, G.C., et al, Electrical Insulation for Rotating Machines, IEEE Press-Wiley, Institute of Electrical and Electronics Engineers, New York, N.Y., 2004.
- Timperley, J., “Specifying Generator Testing,” Proceedings of the International Electric Machines and Drives Conference, Institute of Electrical and Electronics Engineers, New York, N.Y., 2001, pages 139-142.
- Standard for Voltage Endurance Testing of Form Wound Bars and Coils for Hydrogenerators, IEEE Standard 1553-2002, Institute of Electrical and Electronics Engineers, New York, N.Y., 2002.
- Audoli, A., and J.L. Drommi, “Advantages of High Voltage DC Dielectric Tests Compared with AC Tests,” Proceedings of the Electrical Electronics Insulation Conference, Institute of Electrical and Electronics Engineers, New York, N.Y., 1993, pages 661-665.
- Gupta, Bal K., “Use of AC and DC Hipot Tests to Assess Condition of Stator Insulation,” Proceedings of the Electrical Electronics Insulation Conference, Institute of Electrical and Electronics Engineers, New York, N.Y., 1995, pages 605-608.
- Rux, Lori, and W. McDermid, “Assessing the Condition of Hydrogenerator Stator Insulation Using the Ramped High Direct Voltage Test Method,” IEEE Electrical Insulation Magazine, November 2001, pages 27-33.
- Gupta, Bal K., and I.M. Culbert, “Assessment of Insulation Condition in Rotating Machine Stators,” IEEE Transactions on Energy Conversion, Volume 7, No. 3, September 1992, pages 500-508.
- Rux, Lori, and S. Grzybowski, “Evaluation of Delaminated High-Voltage Rotating Machine Stator Winding Groundwall Insulation,” Conference Record of the 2000 IEEE International Symposium on Electrical Insulation, Institute of Electrical and Electronics Engineers, New York, N.Y., 2000, pages 520-523.
- McDermid, W., and B.G. Solomon, “Significance of Defects Found During High Direct-Voltage Ramp Tests,” Proceedings of the Electrical Insulation Conference/ Electrical Manufacturing & Coil Winding Conference, Electrical Manufacturing & Coil Winding Association Inc., Imperial Beach, Calif., 1999, pages 631-636.
Jan Stein, an electrical engineer with EPRI, was the project manager for the survey and analysis work described in the article. Greg Stone, PhD, is a dielectrics engineer with Iris Power LP. Bal Gupta is principal scientist/engineer with AOK Technologies Inc. Greg and Bal performed the research described in the article.
µ Peer Reviewed
This article has been evaluated and edited in accordance with reviews conducted by two or more professionals who have relevant expertise. These peer reviewers judge manuscripts for technical accuracy, usefulness, and overall importance within the hydroelectric industry.
Types of Hipot Tests
There are three basic types of hipot tests: alternating current (AC), very low frequency (VLF), and direct current (DC).
During a 60-Hertz (Hz) AC hipot test, the voltage is increased slowly (in about one minute or less) from zero to the hipot test voltage level, then maintained at that level for one minute. A substantial test transformer is needed to energize the capacitance of a hydro generator stator winding – about 50 to 100 kilovolt-amperes (kVa). The AC hipot test is a pass/fail test.
The VLF hipot test is a variation of the AC test where the frequency of the applied voltage is 0.1 Hz. A VLF test takes advantage of a physically much smaller test set, due to the much lower capacitive current that is needed to energize the stator at 0.1 Hz.
Figure 1: This plot shows test voltage vs. leakage current during a direct current hipot ramp test. The departure from linearity (red line) indicates that insulation problems exist within the winding.
DC hipot tests may either be a conventional hipot test or a controlled rate-of-increase hipot test. The conventional DC hipot test is similar to the AC test. The voltage is increased slowly from zero to the specified hipot voltage, then maintained at that level for one to ten minutes. This is a pass/fail test.
In a controlled DC overvoltage test, the applied voltage is changed in a controlled manner. The voltage may be increased manually in a series of steps or ramped up at a controlled rate of voltage increase (typically 1 or 2 kilovolts per minute) to the maximum test level. The measured current vs. applied voltage is monitored as the test progresses and the voltage increases. Abnormalities or deviations from a straight line increase in the current vs. voltage plot (see Figure 1 above) may indicate insulation problems. The test provides diagnostic information regarding the present condition of the stator winding insulation and serves as a proof test, if the insulation withstands the hipot test voltage.