High-voltage insulation testing: why and how?


The benefits offered by using high-voltage insulation testing as a diagnostic tool are all too often neglected, says Mark Palmer of Megger. He explains why it is needed, what it has to offer and how best to carry it out.
ffective and reliable insulation is essential for the correct and safe operation of virtually every item of electrical equipment. Even in low-voltage systems, regular insulation checks are highly desirable and, in some cases, such as portable appliance testing (PAT), they are a legal requirement.
In medium-voltage systems, insulation testing is even more important, because insulation is often under greater electrical stress and failures are likely to be more costly and, potentially, more dangerous.
Equipment manufacturers, of course, routinely perform insulation tests on their products before supplying them to the end user. Why, then, are further tests needed? The first part of the answer is that it is by no means unknown for insulation damage to occur while equipment is being installed or serviced. More important, however, is that even the best insulation degrades over time.
While this is unavoidable – all insulation starts to deteriorate from the moment it’s put into service – well designed equipment, operated within its ratings, should give many years of reliable service. Nevertheless, the ability to predict accurately when that period of reliable service is coming to an end is an invaluable aid to avoiding costly downtime and unplanned maintenance. Testing is the key but it can only be a reliable indicator if the tests are properly performed, using appropriate equipment.
Dependable results
The first factor to be considered is the test voltage. For dependable results, this needs to be high enough to effectively measure the insulation resistance, but not so high as to overstress the insulation during the test. Some standards relating to specific equipment insulation testing, such as IEEE43:2000, recommend the use of voltages greater than 5kV for an insulation resistance test.
For medium-voltage installations, this means that the usual choice is an insulation tester with a nominal operating voltage of 5kV or 10kV. There is a little more to this choice, however, than simply reading the headline voltage of the tester data sheet, as we shall see.
First, we need to examine briefly what happens during an insulation test and how the test equipment reacts.
When the test voltage is applied to a piece of insulation, a current flows. This current is made up of four major components. The first is the capacitive charging current, which is initially large but in very short time decays exponentially to a value close to zero, provided that the instrument has a sufficient current capability to fully charge the capacitance of the item under test.
The second component is the absorption or polarisation current, which is the result of three effects – a general drift of free electrons through the insulator as a response to the applied electric field, molecular distortion caused by the field and alignment of polarised molecules. This current also decays toward zero but over a very much longer timescale than the capacitive current.
Surface leakage
The third current component is surface leakage, which is present because the surface of the insulation is invariably contaminated to a greater or lesser extent. This current is constant with time but is highly dependent on temperature.
The final current component is conduction current, which is the current that would flow through the insulation if it were fully charged, and full absorption had taken place. This current also remains constant with time. Accurately measuring the conduction current, which is often measured together with the surface leakage current, is the prime objective of insulation testing.
The surface leakage current may, however, be excluded from the measurement if the instrument features a guard connection.
Now let’s examine how an insulation tester behaves when a test is initiated. At first, it must supply the relatively large capacitive charging and polarisation currents.
Portable testers
Portable insulation testers, however, often use high-impedance voltage sources, both for safety and to make the size and weight of instruments manageable. This means that their output voltage falls to a fraction of its nominal value while these currents are flowing.
As we have seen, the charging and polarisation currents reduce with time and, as they do so, the output voltage of the tester will rise. Unfortunately, not all testers are created equal. For those with a good load curve, the output voltage rises quickly as the current decreases and soon reaches a plateau close to the nominal voltage rating of the instrument.
For instruments with a poor load curve, the rate at which the voltage rises as the current decreases is much slower, and the start of the plateau region is poorly defined. In many cases, this means that the desired test voltage is never achieved.
As a result, users of such instruments may be led to believe that the test has reached a steady state condition long before the voltage has actually stabilised. The test may, therefore, indicate acceptable insulation performance when, in fact, that performance has only been checked at voltage which is much lower than that which was intended.
Maximum value
When selecting a high-voltage insulation tester, it is, therefore, beneficial to examine the instrument’s load curve, but this is by no means the only consideration. Another important factor, particularly if diagnostic testing is intended, is the maximum value of insulation resistance that can be measured.
This is often a source of confusion, as many insulation testers will give a “greater than” indication when their measurement range is exceeded. Bear in mind, however, that “greater than” is not a measurement; it’s merely an indication that the result is beyond the measuring range of the instrument. In go/no-go testing, it may be sufficient to know that the insulation resistance is greater than 1 gigohm (1,000 megohm), but the situation is very different with diagnostic testing.
Consider, for example, an item of equipment where the insulation resistance has been recorded over a period of years as relatively steady at around 100 gigohm. A new measurement, however, indicates that this has fallen to 40 gigohm and a further measurement taken a period of time later shows a fall to 10 gigohm.
Potential problems
Clearly, these changes show that there is a potential problem that needs investigation before it can develop into a serious fault. If the same tests had been carried out with an instrument that simply indicated infinity for all resistance values above 1 gigohm, no change at all would have been detected and the insulation would have been given a clean bill of health.
This example illustrates not only the importance of choosing instruments with extended measuring ranges but also of performing insulation tests on a regular basis and recording the results. The database so produced is a powerful tool for initiating preventative maintenance and eliminating the cost and inconvenience of breakdowns.
In this respect, it is worth noting that many of the latest insulation testers have internal facilities for storing test results. These can then be downloaded to a PC for analysis and archiving, a procedure that not only saves time but also eliminates the risk of transcription errors.
One final factor that should be considered when choosing a high-voltage insulation tester is whether a selectable breakdown detector is needed.
With such a detector enabled, the test can be terminated immediately, before a breakdown of insulation and possible damage. If, however, it is an advantage to allow a breakdown to occur – for example, to assist in the location of the weakness in the insulation – the ability to disable the detector is an advantage, as is an instrument with a good short-circuit current capacity.
High-voltage insulation testing, particularly when it is carried out on a regular basis, is an invaluable diagnostic tool and an important aid to predicting and preventing equipment failures.
Maximum benefits can only be achieved, however, if the insulation tester is well chosen. Essentials are a good load curve, an extended measuring range and, for true diagnostic capability, a full battery of pre-programmed tests, including polarisation index, dielectric absorption ratio, step voltage and dielectric discharge. Instruments, like those in the Megger range, which meet all of these requirements, will quickly repay their comparatively modest purchase price.