There are a lot of myths and misconceptions about Hipot testing and destructive testing in general. Hopefully this article can shed some light on these misconceptions.
This article is not written to support the generic use of hipot testing but to “de-criminalise” its use when properly applied with understanding of the principles.
DESCRIPTION OF THE TEST
A high-potential (hipot/withstand/pressure/over-voltage) test is a test that is designed to provide insulation with a higher electric field stress than under normal operation, with the intention that the compromised part of the insulation breaks-down (fails).
The test voltages are chosen to be far below the breakdown voltage of the insulation. During breakdown (also partial breakdown) irreversible and cumulative changes occur that degrade insulation and provided this does not happen then the insulation is largely unaffected.
The failure-on-test is more tolerable than an in-service failure especially for new installations.
The limited energies of the test result in lower levels of collateral damage which normally result in a quicker and cheaper repair.
The failure can then be located and the damaged part of the insulation can be repaired. The repaired insulation contributes towards a more reliable system with reduced risk of failure than were the test not performed.
A hipot test is a destructive test BY DEFINITION irrespective of the voltage source that is used.
Hipot tests do not require any measurement to accompany the test. The result is either a pass if the insulation does not break down or a fail if the insulation does break down.
The hipot test has two control variables : The voltage source and the duration of the test.
The voltage may be ac or DC and, if ac then may be of many different frequencies.
The test voltage should be chosen low enough so as not to degrade the condition of healthy parts of the insulation system but high enough to accelerate any existing damage within the insulation at a sufficient rate as to likely cause breakdown during the test time.
The test time should be chosen in conjunction with the test voltage to give sufficient time for most types of damage to convert to localised failure of the insulation. Increasing the test time has a diminishing return so the test time should also be chosen with efficiency of testing in mind.
There are occasions where failure of the insulation is not tolerable and therefore a hipot test should be avoided. There are a number of alternative diagnostic tests that can be performed at around the operating voltage or below in order to assess the condition of the insulation.
A hipot test is not infallible – insulation systems that pass hipot tests can and will fail in the future. The test argument is not about individual cases but about failure rates. Tests that reduce failure rates (or correspondingly increase reliability) are good tests. test that increase failure rates are counter-productive and test that have no effect on the failure rates are of no value.
A diagnostic can be monitored during the performance of the hipot test. The more sensitive the diagnostic, the better the monitoring (for example using Tan Delta instead of leakage current). If the monitored characteristic remains stable and good then the test times can be reduced but if the diagnostic changes during the measurement then the test time can be increased to allow the proceeses to complete.
In addition diagnostic measurements can be made prior to the hipot test to defer or eliminate subsequent hipot testing for insulation systems where failure is not desirable. There are risks that the switching transients can stress the insulation in any event but these are short duration events and likely not as aggressive as long-duration withstand tests.
1. The hipot test time is shortened to reduce the possibility of failure during test or in an attempt to optimise the test by reducing the time taken.
This is entirely counter-productive and likely to increase the risk of failure post-test. The worst-case scenario is that insulation damage is converted into an electrical tree by the test and the tree grows but the insulation does not fail on test because of insufficient time. Upon return to service the insulation fails and the hipot test is blamed. In fact the hipot test had done its job but wasnt given the time to complete the process.
2. Voltages are borrowed from other standards and other voltage sources.
Test procedures that use “twice the operating voltage plus 1kV” or similar “rules of thumb” cannot be applied across different devices (eg motors and cables) or transported from dc tests to ac tests. You risk providing to much stress or too little stress on the system.
3. The hipot test is used with a diagnostic interpretation.
Insulation that fails a hipot test has likely suffered further cummulative degradation at the point of failure. It is unwise to return a system to service in this event and it is fallacious to assume that the insulation is healthy because the device continues to operate at normal voltages. A hipot test has only two outcomes : Pass or Fail and if it fails you cannot interpret that diagnostically.
4. The hipot test is preceded and succeeded by an Ir test.
I assume that this process (which is quite common) was used in order to quantify the effect of the hipot test on the insulation. Unfortunately the Ir data is so coarse that decisions are most often not possible. The full range of possibilities is evident. Ir increases after test, stays the same or decreases. The decrease can be small, moderate, large or massive (eg >1Tohm to 500Mohm) and still not reliable conclusion drawn.
My feeling is that this is a waste of time.
5. A soak test is the same as a hipot test.
A soak test is designed to eliminate infant mortality. A hipot test will also detect infant mortalities but the higher voltages are designed to accelerate a wider range of insulation degradation.
Failure on test with a soak test can be spectacular with high levels of collateral damage un;like when tested with limited power sources.
1. Hipot testing is damaging to healthy insulation.
2. DC hipot testing is less damaging than ac hipot testing.
2. The lower final charging current of a DC test means that it is “less destructive”.