Gone are the days of old, when the traditional DC (Direct Voltage/Current) hipot was the recommended method for testing these modern cables that form the critical backbone of the distribution electrical power network. Extensive research, often rooted in the hard lessons learned through negative experiences in the field, has brought about changes to and improvements in the approaches and methods used when testing MV and HV cables.
The traditional high voltage “DC Hipot” and/or the routine insulation resistance measurement may still be a viable and useful technique for testing various types of electrical apparatus. However, when it comes to shielded MV and HV cables, this approach has been found to be seriously lacking and, in some cases, potentially damaging to the cables themselves. The latest IEEE400 clearly states that high voltage DC should not be used on extruded cables. Critically significant research around the world and focused studies like the CDFI project led by the National Electric Energy Testing Research and Applications Center (NEETRAC) at Georgia Tech has allowed International Standard bodies, such as the IEEE, to lead the way in prescribing better testing and diagnostic methods and guidelines for detecting abnormalities in MV and HV cables and/or their accessories.
AC (Alternating Voltage/Current) has effectively replaced the DC method for testing these cables. While technically beneficial to the test engineer in evaluating the true condition of these cables, this change also introduced some practical challenges to the test equipment and users involved. This is mainly because shielded MV and HV Cables are effectively very large capacitors, capable of storing large amounts of electrical charge – their capacitances getting larger with the length of the cable.
Energizing the cable’s capacitor, while relatively easy with DC, is not as simple with AC due to the excessive and continuous reactive power demands AC requires. These power demands increase with cable length, frequency of AC and applied voltage. See formula here.
For example, at power frequency 50/60Hz, all you need is a short piece of high voltage cable to overload and trip a typical 15A 120V AC breaker feeding a piece of AC 60/50Hz HV test equipment. However, with the same piece of AC 50/60Hz HV test equipment, you could easily test a very large power transformer or piece of switchgear without any danger of overloading the 120V AC supply. Clearly demonstrated and plainly said, cables are different.
For this reason, the frequency of the AC waveform was lowered to reduce this excessive power demand to a more practical value, nominally 0.1Hz – also called VLF (Very Low Frequency). This means that the reactive capacitive power consumed by a cable is 600 times less at 0.1Hz compared to 60Hz. This can have profound practical implications in the field.
A very good practical standard IEEE400 and its point document IEEE400.2 spell out the voltage test levels and durations to be applied in the field to these cables. Although VLF is low frequency and generally a Sinewave, VLF is not DC and does not have any of the potential negative side effects that DC may have on the cable. In addition, VLF is a lot more effective in evaluating a cable’s true condition. VLF is now the most widely used method for performing withstand AC testing on MV cable in the field and is also used as the power supply of choice for performing TD (Tan Delta) and PD (Partial Discharge) advanced cable diagnostics.
Tan Delta and its close cousin, Power Factor, have been used for many years to detect insulation integrity of electrical apparatus, such as transformers, insulation oil, switchgear, etc. Measuring TD at VLF allows cable owners to effectively assess the condition of both new and established cables. Unlike insulation resistance measurements, Tan Delta is independent of length of cable, making it very useful for comparative measurements and evaluating measurement data.
A major cause for cable failures in the field has been the ingress of water into the cable system. In the case of XLPE insulated cables, this can cause the cancer-like growths of thousands of water trees in the insulation. Over time these trees can weaken, and in many cases, lead to complete failure of the cable insulation.
Measuring TD at VLF has been found significantly efficient in detecting water trees and water degradation in cables. The IEEE400.2 has clearly defined acceptance criteria for the assessment of the cable insulation system, allowing the test operator to easily interpret the results of the tests. As with power factor, good TD results are low and stable while poor TD tests show large unstable TD values with large tip-up’s (gradient) of TD versus applied voltage. See Figure 1.
Traditionally, TD was done using an external TD measurement apparatus that was placed between the VLF source and the cable under test. With the latest VLF instruments, the TD measurement circuitry is sometimes integrated into one compact, very portable test instrument.
See Figure 2.
In addition to TD, the VLF AC waveform also allows users to detect and locate Partial Discharge (PD) electrical activity in the cable system. Partial Discharge is defined as a partial (not complete) breakdown of the electrical insulation that, over time, can result in complete failure of that insulation. This is very useful for both acceptance testing of new cables and maintenance testing of existing installations, particularly when poor workmanship is involved. For Partial Discharge to occur, you need an AC voltage. Unlike DC, both AC VLF and AC Power Frequency allow for the detection of potential PD in a cable system. Using advanced signal processing, these fast PD transients can be measured, and their origin located in a cable using TDR (Time Domain Reflectometry) and other methods. See Figure 3. While this method often requires a well-trained user, what was once only done by highly skilled operators in special shielded cable test bays, can now be routinely done in the field.
Very Low Frequency, Tan Delta and Partial Discharge are being routinely performed on MV cable installations. In recent years, the advent of newer technologies and improvements in materials has now made available VLF test systems with both TD and PD capabilities that can go up to 200kV to test transmission class cables. Typically, utilities would often just energize a transmission class cable and hope for the best. If the cable had a pending fault, the resultant failure would often be catastrophic since the fault currents would be excessive, often with collateral damage to other equipment. More cautious utilities would employ testing companies, normally at great expense, to perform offline 50/60Hz tests with very large test equipment. These newer VLF test systems allow the cable owners to more affordably test these cable systems with smaller test equipment to ensure that the cables are fit for service before being connected to the electrical grid. See Figure 4.