How to perform AC withstand voltage tests on high-voltage cross-linked cables in the field using a series resonant device? To answer this question, we must first understand the cable structure. Currently, both internationally and domestically, XLPE (cross-linked polyethylene) insulated power cables are increasingly replacing the original oil-filled paper insulated power cables. However, due to the large test capacity and limitations of testing equipment, the power system has continued to use DC withstand voltage testing methods for decades in its pre-commissioning testing of cross-linked cables.
In recent years, research findings from numerous international and domestic research institutions have indicated that DC testing causes varying degrees of damage to XLPE cross-linked polyethylene cables. Some studies suggest that the XLPE structure has the ability to store and accumulate unipolar residual charge. If this residual charge cannot be effectively released after a DC test, the addition of an AC voltage peak to the residual charge after commissioning may cause the cable to break down.
Some domestic research institutions believe that during DC withstand voltage tests of cross-linked polyethylene (XLPE) cables, due to the space charge effect, the actual electric field strength in the insulation can be up to 11 times higher than the working electric field strength of the cable insulation. Even if an XLPE insulated cable passes the DC test without breakdown, it will still suffer severe insulation damage. Secondly, because the distribution of the applied DC voltage field strength differs from that of the operating AC voltage field strength, DC tests cannot realistically simulate the overvoltages the cable experiences under operating conditions, nor can they effectively detect defects in the cable itself, cable joints, and construction processes. Therefore, the use of non-DC methods for withstand voltage testing of XLPE cables is receiving increasing attention.
Currently, ultra-low frequency (VLF) power supplies are used for withstand voltage testing of medium and low voltage cables abroad. However, due to the low voltage level of these VLFs, they cannot be used for testing high-voltage cables of 110kV and above. In China, this method has also been used for low-voltage cables, but due to limitations in testing equipment, it has not been widely adopted. In recent years, with the construction and renovation of urban and rural power grids, XLPE cross-linked cables are becoming increasingly common. Some cables are put into operation after only a DC withstand voltage test, leading to frequent instances of cable or cable head breakdown under operating voltage. Therefore, new testing methods are being explored.
Discussion on testing standards
Due to limitations in equipment capacity and size, there are currently no national standards for conducting AC withstand voltage tests on high-voltage power cables after installation. However, questions have also arisen regarding the standards for DC withstand voltage tests for the reasons mentioned above. The CIGRI International Large Electric Systems Working Group 21's "Recommended Guidelines for Acceptance Testing of High-Voltage Extruded Insulated Cables" raises concerns about the currently used DC withstand voltage test methods and recommends using AC test methods at or near the power frequency (30-300Hz).
The IEC 60840 standard, in its testing standards for cables laid at 45-150kV, added an AC testing standard of 1.7U05 minutes or 1U024 hours to the original DC testing standard. However, in the IEC 62067/CD draft for the 220kV level, the DC testing standard for cables laid after installation was removed, leaving only the AC testing requirement: 1.4U060 minutes for 20-300Hz. To more effectively conduct acceptance testing on cross-linked cables after construction, the North China Electric Power Group Corporation added an AC withstand voltage test standard to the cable main insulation withstand voltage test section in its revised "Code for Acceptance and Preventive Testing of Power Equipment".
Subsequently, many places in China also introduced their own local testing standards, with testing frequencies mostly between 30-300Hz. The test voltage for medium and low voltage cables is 1.6-2.0 times the phase voltage, and the test voltage for high voltage cables is generally 1.4-1.7 times the phase voltage. The specific values vary slightly from place to place. Zhejiang recommends a test frequency of 45-65Hz for medium and low voltage cables and 35-75Hz for high voltage cables.
Cable testing frequency controversy
Due to the large capacitance of cables, traditional power frequency test transformers are bulky and heavy, and high-current power supplies are not readily available on-site. Therefore, series resonant AC withstand voltage test equipment is generally used. This significantly reduces the input power capacity and weight, making it easier to use and transport. Initially, inductive series resonant equipment (50Hz) was commonly used, but it suffered from poor automation and high noise levels. Therefore, the TPXZB series frequency-modulated (30-300Hz) series resonant test equipment is now widely adopted, achieving higher quality numbers (Q values) and offering advantages such as automatic tuning, multiple protections, low noise, and flexible configuration options (significantly reducing unit weight).
Based on a comprehensive review of relevant technical data both domestically and internationally, selecting an appropriate test frequency range is a crucial issue. There are several different viewpoints and approaches in this regard. Based on current domestic and international perspectives, we can summarize them into three categories: Category 1: a relatively wide frequency range of 30-300Hz, 20-300Hz, and 1-300Hz; Category 2: the power frequency range of 45-65Hz and 45-55Hz; and Category 3: close to the power frequency range of 35-75Hz.
(1) Category 1 has a wider frequency range
The test guidelines published by Working Groups 21 and 09 of the International Conference on Large Electric Systems recommend a frequency range of 30-300 Hz. However, in practice, lower frequencies also have good equivalence. The draft standards IEC 60840 and IEC 62067 (2001 and 2000) stipulate that 20-300 Hz can be used.
Some foreign manufacturers design series resonant reactors, and in special cases, they may use a minimum frequency of 25Hz or 20Hz. Of course, the lower the frequency, the longer the tested cable (capacitance) can be increased. However, this also enlarges the reactor core, increasing its weight. Some data indicate that AC tests in the 1-300Hz range are equivalent to power frequency AC tests, suggesting that in practical applications, the lower frequency limit can be even lower, for example, less than 20Hz or even down to 0.1Hz. This further demonstrates that within such a frequency range, the voltage distribution and dielectric characteristics of the various media within the insulation remain essentially the same.
Is it appropriate to operate at a frequency exceeding 300Hz? Some reports suggest that as the frequency increases, the losses of the series resonant reactor and the excitation transformer decrease. However, the polarization heating of the capacitor dielectric of the test sample must be considered. Therefore, a frequency higher than 300Hz is not advisable.
(2) Category 2 is the power frequency range
Internationally, industrial frequencies mainly refer to 50Hz and 60Hz. Therefore, the IEC standard specifies that the industrial test frequency range for high-voltage insulation is 45-65Hz, while the rated power frequency in my country is 50Hz. GB/T16927.1-1997 specifies that the power frequency test frequency range is 45-55Hz.
The view that the test voltage for power frequency cables must also be power frequency is a rather conservative one. Regarding this issue, it should be emphasized that the purpose of acceptance and preventative testing is determined by the ability to detect insulation defects. As long as the voltage distribution within the insulation medium is the same at different frequencies, and the ability to detect insulation faults is essentially the same, the test objective can be achieved. Therefore, even using a frequency with a wider range than the power frequency is acceptable.
In the mid-1990s, extensive and meticulous fundamental research was conducted to select an appropriate frequency range for AC withstand voltage tests. It was found that within the frequency range of 30-300Hz, the breakdown characteristics of several typical insulation defects inside rubber and plastic cables showed no significant difference. This finding is considered reliable and has been widely adopted. Analysis suggests that the main reason for these favorable breakdown characteristics at different frequencies is the excellent coaxial insulation structure, the single insulating medium, the relatively pure material, and the reasonable and regular electric field distribution. Therefore, the voltage distribution within the structure is the same at different frequencies, creating conditions for a wide frequency range test.
Oil-paper insulated cables have always been tested for withstand voltage using a DC voltage with a frequency of zero, and the results have been excellent and have not been questioned for decades.
(3) Category 3 is close to the power frequency, 35-75Hz
Abroad, breakdown tests were conducted on samples of normal XLPE (cross-linked polyethylene) insulated cables at different frequencies. The results showed that the breakdown voltage fell within a 95% confidence level across frequencies of 35-75Hz. Therefore, some argue that the optimal test voltage frequency is 35-75Hz, which is also close to the operating voltage frequency of 50Hz. It is important to note that the above test results pertain to breakdown tests of normal insulation. The test voltage values used in acceptance and preventative tests are lower; they can only break down defective insulation weaknesses (mechanical damage, water trees, stress cones in terminations or junction boxes, incorrect construction or material selection, etc.), and are completely insufficient to break down the normal insulation of the cable itself. It is evident that the purposes and working mechanisms of the two tests are different. It seems unnecessary to "extend" the breakdown characteristics of normal insulation at 35-75Hz to the detection of insulation defects.
Selection of series resonant devices
Cables with voltage levels of 35kV and below are used in large quantities, and the workload of testing is large. Therefore, such withstand voltage test equipment should be small in size and light in weight. The cable series resonant test equipment produced by Yangzhou Top was born out of this need.
In this device, the weight of components that a single person can move on-site does not exceed 30kg, and the weight of components that two people can move on-site does not exceed 60kg. It is suitable for on-site handling. The reactor section uses dry epoxy casting, which is aesthetically pleasing and reliable, and suitable for various types of cables. This device has the following characteristics:
(1) Easy to operate
It features automatic/manual tuning functions; a large screen displaying parameters such as voltage, current, and frequency; overvoltage and overcurrent setting and protection functions; detuning protection function; automatic timing and time-out shutdown function; and zero-point start function for voltage regulation to ensure tuning safety and zero-start voltage boost in the high-voltage circuit.
(2) Flexible configuration
The reactors are multi-unit type, and the tested equipment can be connected in series or used together. Moreover, the number of reactors brought to the field area can be determined according to the tested equipment.
(3) Lightweight and small size
Taking a 2km long 10kV cable (high voltage current 2.5A) as an example, the weight of a single reactor is 35kg (2 units), the weight of the frequency converter control box is 12kg, and the weight of the excitation transformer is 15kg.
