Introduction Metal oxide surge arresters are widely used in power systems. While the vast majority operate well, explosions during operation do occur. Analysis of the causes of these explosions reveals a common problem: quality issues with the arrester itself. These include poor valve plate performance and unreasonable parameter design, inadequate internal insulation materials, and poor assembly processes leading to moisture absorption during operation. Accidents caused by poor insulation account for a relatively large proportion. Therefore, in addition to careful product selection, rigorous acceptance testing and regular preventative testing are essential to identify and address problems promptly and prevent accidents. 1. Preventative Testing The preventative testing of metal oxide surge arresters mainly involves measuring the critical operating voltage U1mA at 1mA (DC) and the leakage current at 75% U1mA DC to check for moisture absorption of the valve plates. This test is susceptible to external factors, which can lead to failures. If this problem is not handled properly, on the one hand, good equipment may be treated as defective, wasting manpower and resources and prolonging the power outage time; on the other hand, it may also overlook the actual faults in the equipment, posing a hidden danger to the safe operation of the equipment. The voltage at 1mA DC refers to the voltage value across the surge arrester when a DC current of 1mA flows through it. In the test, a DC voltage is applied to the surge arrester, and the voltage increase is immediately stopped when the DC current reaches 1mA. The voltage U1mA of the surge arrester is then quickly read, and then the voltage is reduced to 75% of U1mA to read the current value flowing through the surge arrester. The schematic diagram is shown in Figure 1. [IMG=Schematic Diagram]/uploadpic/THESIS/2007/12/2007122516281890663O.jpg[/IMG] The measured value of U1mA should not differ from the initial or manufacturing value by more than 5%. Excessive U1mA reduces the insulation margin of the electrical equipment protected by the surge arrester, while excessively low U1mA can cause the surge arrester to explode under transient overvoltages during various operations and faults. Regulations stipulate that the DC leakage current at 75% U1mA should not exceed 50mA. Generally, the main reasons for unqualified insulation test data of surge arresters are considered to be the following: High air humidity; Insufficient angle between the high-voltage lead and the equipment during testing; Internal moisture in the surge arrester, mainly due to poor assembly processes leading to moisture absorption during operation; Incomplete disconnection of equipment connected to the surge arrester during testing. 2. Field Tests On August 1, 2005, during the surge arrester test of the 35kV 383 line at our company's 220kV Lianshui substation, it was found that the DC 1mA voltage values ​​of the three-phase surge arresters were all too low, as shown in Table 1. [IMG=Field Tests]/uploadpic/THESIS/2007/12/20071225162826489708.jpg[/IMG] On November 29, 2005, during the surge arrester test of the 110kV busbar at our company's 110kV Gaogou substation, it was found that the DC leakage current of the C-phase surge arrester at 75% U1mA was relatively large, and the test results are shown in Table 2. [IMG=Test Result Current Comparison]/uploadpic/THESIS/2007/12/2007122516283278633M.jpg[/IMG] 3. Cause Analysis In the test on August 1, 2005, the humidity was high on the day of the test, making it difficult to reach below 80% during measurement. Furthermore, there was a water film on the surface of the surge arrester. Based on the above measurement data, the three-phase insulation resistance decreased significantly after the humidity increased. During the test, this group of surge arresters was not connected to any equipment. Therefore, the possible reasons for the unqualified measurement results are as follows: High air humidity, dirty surface of the surge arrester; Moisture inside the surge arrester. In the test on November 29, 2005, due to the technical renovation work of the high-voltage transformer, the surge arresters were brand new and not connected to any equipment, so the problem of moisture was unlikely. However, due to the messy site, the angle between the high-voltage lead and the tested equipment was difficult to control, remaining less than 30°. This may be the reason for the relatively large error in the test data. The test results are shown in Table 3. [IMG=Data Error Comparison]/uploadpic/THESIS/2007/12/2007122516283741005J.jpg[/IMG] 4. Cause Confirmation The reason why high air humidity and a dirty surface on the surge arrester lead to a lower DC 1mA voltage value is easy to understand. When a DC voltage is applied across the surge arrester, the current passing through the arrester has two parts: one is the current through the internal valve plates of the arrester, which is related to the characteristics of the arrester; the other is the leakage current through the surface of the arrester. If the air humidity is high, the leakage current on the surface of the arrester will be relatively large, so the total current during voltage boosting can easily reach 1mA, making the voltage lower. That is, U=RI, where R is the insulation resistance and I is 1mA. When the air humidity is high and the surface of the surge arrester is dirty, R is smaller, and U is also smaller. The influence of the angle between the high-voltage lead and the surge arrester on the test data can be demonstrated by testing a group of 110kV surge arresters, as shown in Table 4. [IMG=110 kV surge arrester test table]/uploadpic/THESIS/2007/12/2007122516304135716F.jpg[/IMG] It can be seen from this that the angle between the high-voltage lead and the surge arrester has a significant impact on the leakage current under 75% DC 1mA voltage. Through the analysis of the above test data, we can also see that: (1) High air humidity and surface dirt will lead to a lower DC 1mA voltage of the surge arrester, but have little effect on the leakage current under 75% DC 1mA voltage. (2) During measurement, the angle between the high-voltage lead and the surge arrester has a significant impact on the leakage current under 75% DC 1mA voltage, but has little effect on the DC 1mA voltage of the surge arrester. Therefore, the following countermeasures were formulated: Clean the surface of the surge arrester; expose the test sample to sunlight or heat it to dry the insulating porcelain bushing; keep the high-voltage line as horizontal and far away as possible, avoiding contact with the surface of the surge arrester's porcelain bushing; disconnect all equipment connected to the surge arrester under test to ensure that the surge arrester can withstand voltage independently. 5. Implementation of Countermeasures On August 1, 2005, at noon, the weather improved, and we conducted a retest based on the above analysis. First, we cleaned the insulating porcelain bushing and dried the surface of the surge arrester with a hairdryer. The main insulation resistance increased to 3000 Mw, showing a significant increase. Table 5 shows the measurement data for the three-phase surge arrester. [IMG=Measurement Data of Three-Phase Surge Arresters]/uploadpic/THESIS/2007/12/2007122516305216650Y.jpg[/IMG] On November 29, 2005, during the retesting of the C-phase surge arrester of the 110 kV busbar of the 110 kV Gaogou substation, we used nylon ropes to pull the high-voltage leads to an angle basically perpendicular to the surge arrester and conducted a pressure test. The results are shown in Table 6. [IMG=Pressure Test]/uploadpic/THESIS/2007/12/2007122516305730240X.jpg[/IMG] As can be seen from the above measurement data, by implementing the established countermeasures, the previously unqualified test data are now within the acceptable range specified in the regulations. 6. Conclusion When measuring the critical operating voltage U1mA at 1mA (DC) and the leakage current at 75% U1mA DC of a metal oxide surge arrester, the following two scenarios should be considered. If one data point is non-compliant, the established countermeasures should be considered first, and the equipment should not be immediately and arbitrarily concluded to be defective. If both data points are non-compliant, the surge arrester is more likely to be defective.