1. Introduction Cross-linked polyethylene (XLPE) power cable insulation has a very high volume resistivity, exceeding 10¹⁷ Ω·m. Under a DC electric field, space charge is easily generated and accumulated, causing distortion of the electric field at local defects in the XLPE medium (such as unavoidable air gaps, impurities, or water trees generated during operation). The local electric field strength increases dramatically by more than 10 times, approximately 30 kV/mm, far exceeding the breakdown field strength of the XLPE medium, leading to local breakdown, irreversible early dendritic degradation, and even breakdown failure. On the other hand, when the DC electric field is removed, the space charge already formed in the medium cannot leak out in a short period due to the high resistance of the medium, forming an additional electric field in the local area. When this additional electric field superimposes with the external operating frequency electric field to form a very high local electric field, it may rapidly break down the XLPE medium. These phenomena frequently occur during DC withstand voltage tests of XLPE power cables; for example, cable lines that pass the DC withstand voltage test may experience breakdown failure shortly after normal power supply. In the early 1980s, it was discovered that under the influence of a DC electric field, the additional electric field effect of space charge in XLPE power cables strengthens the electric field at the tips of water trees, triggering partial discharge of the dielectric. This releases a large number of high-energy charged particles that continuously bombard the dielectric molecular chain segments at the ends of the water trees and the walls of the water tree channels, causing the dielectric molecular chain segments to break and degrade. The water trees rapidly transform into electrical trees, accelerating the early deterioration of the insulation performance of XLPE power cables [1] and greatly shortening the service life of the cables. Some developed countries with large cable usage have eliminated the DC withstand voltage test in the preventive testing of XLPE power cables and successively introduced oscillating wave voltage test, 0.1 Hz ultra-low frequency voltage test, and power frequency voltage test methods [2-5]. In the late 1990s, my country revised the provisions on preventive test methods for cables in the preventive test regulations for electrical equipment and no longer recommended the DC withstand voltage test. As early as the mid-1980s, countries such as Germany, Austria, the United States, and Japan began to use ultra-low frequency (0.1 Hz) withstand voltage testing as a non-destructive testing method to detect insulation defects in operating XLPE power cables, and carried out a large number of laboratory and field test studies. Research reports from the German Power Plant Association (VDEW) and Working Group 21.09 of the International Conference on Large Electric Systems (CIGREWorking Group 21.09) indicate that ultra-low frequency (0.1 Hz) withstand voltage testing is still recommended for testing medium-voltage XLPE insulated power cables. On the one hand, this test does not accumulate space charge or distort the local electric field in the XLPE insulation of the cable; on the other hand, it can detect defects such as water treeing aging in the cable insulation at a relatively low test voltage, thus causing less damage to the XLPE insulation [2-5]. Since 1996, some cities in my country, such as Shanghai, Yantai, and Beijing, have gradually adopted ultra-low frequency (0.1 Hz) withstand voltage testing combined with dielectric loss (tgδ) measurement of cable insulation under ultra-low frequency (0.1 Hz) voltage as a non-destructive testing method to detect insulation defects in cables. At the same time, research on the effectiveness of ultra-low frequency (0.1 Hz) withstand voltage testing has been carried out. There are many reports in domestic and foreign literature on the results of extended research on partial discharge testing or dielectric loss testing under ultra-low frequency (0.1 Hz) power supply. However, there are few research results on the effectiveness of ultra-low frequency (0.1 Hz) withstand voltage testing and the equivalence between ultra-low frequency (0.1 Hz) withstand voltage testing and power frequency withstand voltage testing. These are also the focus of great attention and heated debate [6]. This paper studies the effectiveness and feasibility of these three test methods as early detection and identification of potential accidents in the operation of XLPE power cables by conducting parallel comparative tests on XLPE power cable test samples with man-made insulation defects using power frequency voltage, oscillating wave voltage, and ultra-low frequency voltage. [b]2 Experimental Research[/b] The parallel comparison test method study was conducted based on the following two conditions: ① The XLPE power cable has manufacturing quality defects and construction quality defects; ② After a period of operation, the XLPE power cable will expose the early aging defects of water trees. The cable defects were discovered and identified by comparing the same defective cable specimens with power frequency voltage, 0.1 Hz ultra-low frequency voltage and 5-6 kHz oscillating wave voltage tests. 2.1 Specimen Preparation Defects were artificially created on YJV-8.7/15 3×185 cables to simulate insulation quality defects in the cable during manufacturing, construction and operation. YJV22-8.7/15 3×185 cables with serious defects that had been in operation for 12 years were also taken from the operation site as specimens, as shown in Table 1. 2.2 Testing Before the tests, the outer sheaths of samples S1, S2, S3, and S4 were removed, and each was divided into three identical samples. Each sample had pre-made silicone rubber terminals at both ends, and these were numbered as follows: S1a, S1b, S1c; S2a, S2b, S2c; ..., S4a, S4b, S4c, for a total of 12 samples. Similarly, the outer sheath of sample S5 was removed, and it was divided into S5a, S5b, ..., S5l, for a total of 12 samples. The three samples of the same number were subjected to power frequency voltage breakdown tests, 0.1 Hz ultra-low frequency voltage breakdown tests, and 5–6 kHz oscillating wave voltage breakdown tests, respectively. After each breakdown, a dummy connector was made at the breakdown point, and the same voltage waveform breakdown test was performed again. The test circuits for the ultra-low frequency method and the oscillating wave method are shown in Figure 1. In Figure 1, Rp is the protective resistor; R1 and R2 form a resistor voltage divider; CRT is an oscilloscope; R3 is a damping resistor; L is an adjustable inductor; and S.G is the ignition ball gap. Based on the capacitance Cx of the test cable and the value of the adjustable inductor L, the frequency f and damping coefficient α of the oscillating wave are calculated using the formulas f = 1/2π√LCx and α = R3/2√LCx. 2.3 Test Results The breakdown voltages of the three test methods are shown in Tables 2, 3, and 4. [b]3 Analysis and Discussion[/b] The distortion of the space charge in the DC electric field by the local electric field of the insulating medium is an indisputable fact. When the total space charge q is concentrated at point x in the medium, it conforms to the Poisson equation: Ect - E = q / (ε0ε), where Ect is the local field strength at point x in the medium, and E is the field strength of the neutral body in the medium. The space charge q can be obtained from the space charge concentration rate dq/dt = γηE, where γ is the conductivity of the medium; and η is the fraction of charge transferred before the medium discharges. When Ect exceeds the local breakdown field strength of the dielectric, the insulation performance of the dielectric will significantly decrease. From the test results in Table 2, it can be seen that the breakdown voltage of the dielectric under alternating electric fields is greatly affected by defects. The power frequency breakdown voltage of the defective cable samples drops significantly, only 30% to 60% of the breakdown voltage of normal cables, mainly exhibiting electromechanical stress breakdown in polymer dielectrics. Let the elastic modulus of the dielectric be Y. When an alternating electric field E is applied to the dielectric, the Maxwell stress f in the dielectric is: f = ε0εE²/2. When the deformation Δh of the dielectric caused by the electric field stress f exceeds the elastic range of the dielectric, the dielectric undergoes mechanical fatigue until electromechanical breakdown. The critical condition for electromechanical breakdown is: EB∝[Y/（ε0ε）]¹/², that is, the critical condition EB for dielectric breakdown is proportional to the elastic modulus Y of the dielectric. Since the elastic modulus of XLPE dielectric is relatively low, it can be inferred that a lower power frequency voltage can effectively detect defects in the insulating dielectric. Generally, it is not easy to determine whether XLPE power cables are damp or have early water treeing defects using short-term tests. In the past, the magnitude of DC leakage current and the three-phase balance coefficient were used for judgment. However, the surface leakage current is often much greater than the polarization current and conductivity current of the insulation medium. Therefore, it often masks the true state of the cable insulation and fails to detect insulation defects in time. The test results in Tables 3 and 4 show that the artificial defects in samples S2 and S4 simulate the state of the cable being damp and having water trees during actual operation. Since the cable terminal was installed after the moisture treatment, the surface leakage current is small. At this time, when an ultra-low frequency voltage is applied, the polarization process of water molecules in the medium can keep up with the change of the applied electric field. The conductivity of the medium is affected by the synergistic effect of impurity ion conductivity, moisture polarization, dielectric electronic conductivity, electromechanical stress and other factors. Macroscopically, it shows a low breakdown voltage. Its breakdown field strength Ec approximately follows the variation law of the combined electron theory: ln Ec = constant + ΔV / (2kT), where ΔV is the width of the excited state of the impurity energy level in the dielectric energy band; k is the Boltzmann constant; and T is the absolute temperature. Similarly, when a high-frequency oscillating voltage is applied, the polarization process of water molecules in the dielectric cannot keep up with the change in electric field. The dielectric is mainly affected by the electronic conductivity factor, resulting in a higher breakdown voltage. Samples S1 and S3 mainly simulate impurities and air gaps in XLPE dielectric, as well as insulation damage during construction. They can be approximated as solid-solid or solid-gas composite dielectrics. When the frequency of the applied electric field is high, the electric field inside the composite dielectric will be distributed according to its dielectric constant. The electric field is higher at the air gaps and impurities, which first causes partial discharge, leading to local breakdown of the dielectric. The local breakdown voltage is inversely proportional to the frequency of the electric field. When the voltage gradually increases to a level that can sustain partial discharge, the dielectric exhibits macroscopic electrothermal breakdown. Sample S5 is a cable specimen that was decommissioned after 12 years of operation. After the dielectric was sliced, stained, and observed under a microscope, a large number of water trees were found. Defects were artificially created and subjected to DC voltage tests, power frequency voltage tests, 0.1 Hz ultra-low frequency voltage tests, and 5–8 kHz oscillating wave voltage tests to verify the equivalence of the DC breakdown voltage test (K = Ux/Uac). The results are listed in Table 5. In the cable medium, the DC electric field is distributed according to resistivity, while the alternating electric field is distributed according to dielectric constant. Comparing the results in Table 5, the K value for the DC breakdown voltage test varies within a wide range (2.6–4.3) depending on the type of cable defect. This is because the gradient of dielectric constant change in the cable medium is much greater than the gradient of resistivity change; the DC breakdown voltage of the same type of defect medium is more than twice that of the power frequency breakdown voltage. If K = 2.6 is selected for the DC voltage test, cable cut defects, pinhead defects, or tip defects are not easily detected; if K = 4.3 is selected, the cable medium may be damaged due to the injection or accumulation of a large amount of space charge by a very high DC electric field. Therefore, the DC electric field cannot effectively detect defects present in the cable medium. The K-value distribution of the oscillating wave voltage test is relatively uniform (1.1–1.5), indicating that it can comprehensively detect cable dielectric defects and has relatively good equivalence with the power frequency voltage test, especially sensitive to indentation defects (K=1.1), making it one of the acceptance test methods after cable completion. The K-value of the ultra-low frequency voltage test varies in the range of 1.2–2.6 depending on the type of defect, and its equivalence with the power frequency voltage test needs further in-depth research. However, the outstanding feature of this test method is its ability to effectively detect moisture ingress and water tree defects in the cable dielectric (K=1.2), and the most common early insulation performance degradation phenomenon of XLPE insulated cables in actual operation is dendritic aging or moisture ingress in the cable dielectric. Relatively speaking, this method is more suitable as a preventive test method for XLPE insulated cables. The power frequency voltage test is of course the ideal test method (K=1), but due to the large capacitance of XLPE insulated power cables, especially high-voltage cable tests, the power frequency test equipment requires a large capacity, and the equipment is large in size and weight, making it inconvenient for on-site testing. Currently, efforts are being made to obtain high voltages close to the power frequency by using various induction methods or frequency conversion methods to resonate with the capacitance of the cable, in order to reduce the size and weight of the test equipment. Compared with power frequency voltage test equipment, the capacity, weight and volume of oscillating wave voltage or ultra-low frequency voltage test equipment can be significantly reduced. However, so far, the output voltage of ultra-low frequency voltage test equipment is not high and can only be used for power cable testing in power distribution systems, and its equivalence with power frequency voltage still needs further research. Oscillating wave voltage test equipment is relatively easy to implement, and field tests can be completed using existing DC power supplies and reactors with adjustable inductance. The problem that still needs to be solved is the research on its equivalence with power frequency voltage [5]. At the same time, partial discharge measurement systems and dielectric loss measurement systems under 0.1 Hz ultra-low frequency voltage and kHz oscillating wave voltage are also under research, in order to achieve both no damage to cable insulation and easy detection of cable insulation defects, further reducing the size and weight of the equipment and meeting the needs of field testing. [b]4 Conclusions[/b] (1) The power frequency voltage test can comprehensively and realistically discover defects and potential operational faults in XLPE cables. It can be applied to the completion test and preventive test of XLPE power cables, especially the completion test and preventive test of XLPE insulated power cables with voltage levels of 110 kV and above. However, how to reduce the size and weight of the power frequency voltage generator needs further research. (2) The 0.1 Hz ultra-low frequency voltage test can effectively discover the moisture and water tree operation defects of XLPE insulated power cables at lower voltages. It can be used as a preventive test method for XLPE insulated power cables in power distribution systems. (3) The kHz oscillating wave voltage test can effectively discover manufacturing quality defects and construction quality defects of XLPE insulated power cables at lower voltages. It is recommended as a completion test method for XLPE insulated power cables. [b]References:[/b] [1] Luo Junhua, Yuan Chunzhi. Study on the dendritic degradation characteristics of XLPE power cables under DC electric field [J]. High Voltage Engineering, 1993, (1). [2] Katsumi Uchida, et al. Study on detection for the defects of XLPEcable lines [J]. IEEE Transactions on Delivery, 1996, 11(2). [3] Eager GS, et al. High voltage VLF testing of power cables [J]. IEEE Transactions on Delivery, 1997, 12(2). [4] DIN VDE0276-1001, 1995. [5] Working Group 21.09. After laying tests on high voltage extruded insulation cable systems electra [R]. 1997, (173): 33-41. [6] Li Xiangzhong. Progress in cable dielectric loss measurement technology at 0.1Hz [J]. Guangdong Electric Power, 1998, (6).