Abstract: Capacitive voltage transformers are widely used in power systems, but a failure caused by a loose bolt, as seen in this case, is extremely rare. For power equipment manufacturers, even a one in ten thousand defective products can pose a 100% hazard to equipment users. Keywords: Capacitive voltage transformer, fault analysis and handling In mid-March 2001, during a pre-commissioning inspection of the newly installed automatic power transfer (BZT) device at the 110kV Jinyuan substation, our bureau's relay protection personnel discovered no voltage on the standby power supply side. Since this voltage was introduced through a capacitive voltage transformer (CVT) installed on the standby power line side, relay protection and high-voltage testing personnel conducted a series of inspections and tests on the CVT and its secondary circuit. The results revealed a serious fault: the CVT's electromagnetic unit was burned out. Maintenance personnel promptly replaced it, preventing an equipment accident. 1. Equipment Fault Discovery Process Our bureau's Jinyuan substation has two 110kV power lines. Under normal operation, one line supplies power to the main CVT while the other is a CVT standby. To quickly switch on the backup line in the event of a permanent fault in the main power supply line, a BZT device was installed in the 110kV system. As shown in Figure 1, the BZT device is connected to the voltages of the north and south sections of the Jinyuan 110kV busbar and the voltages of the two line sides. Using the device's switching handle, each line can be switched to either the main power supply line or the backup line, and the corresponding busbar voltage, line voltage, and secondary circuit can be switched accordingly. Under normal operating conditions, the Zijin line is the main power supply, and the Tjin line is the backup. In this case, the voltage of the north busbar of the Jinyuan 110kV and the voltage of the Tjin 2 line side are connected to the BZT device. The north busbar voltage reflects the operating status of the main power supply, and the Tjin 2 line side voltage reflects whether the backup power supply is normal and can function as a backup. This BZT device was installed in December 2000. In mid-March of this year, relay protection personnel conducted a pre-commissioning inspection of the device. The staff measured the voltages of the two busbar sections and the voltage of the Zijin line side on the terminal blocks behind the device panel; these were normal. However, when measuring the secondary voltage of the CVT on the Tjin 2 line side of the backup power supply, there was no voltage. At the time, the staff assumed the line was not energized and put the matter aside, only inspecting the device itself. Because all equipment at the station required periodic high-voltage testing, the main power line equipment could only be shut down for testing after the backup line was put into operation. After contacting dispatch, the operators put the T-Gold line into operation to supply the entire station load, while the Zi-Gold line was shut down. At this point, the relay protection personnel confirmed that the T-Gold line was energized and measured the voltage on its line side again on the BZT device panel, but still found no voltage. The CVT secondary fuse was not blown; the circuit was disconnected and the conductor cores were checked, and there were no problems; the secondary fuse was removed, and the voltage was measured directly at the secondary output terminal, but still no voltage. The relay protection personnel then realized that there might be a fault inside the CVT. Therefore, after quickly completing the high-voltage test on the Zi-Gold line equipment, the Zi-Gold line was put into operation, the T-Gold line was shut down, and the primary lead of the CVT on its line side was removed for testing. This capacitive voltage transformer was put into operation in December 2000, and its electrical principle is shown in Figure 2. The high-voltage testing personnel first tested the high-voltage capacitor C1, medium-voltage capacitor C2, and total capacitance of the CVT, and then tested the dielectric loss. The results showed little change compared to the factory and pre-commissioning test data, indicating that the capacitor divider unit was functioning correctly. To determine the problem with the CVT's electromagnetic unit, the personnel first used a multimeter in resistance mode to measure the resistance of the primary coil of the medium-voltage transformer, finding it to be over 500 ohms. Then, they applied an AC voltage to the primary coil of the medium-voltage transformer and measured the secondary voltage; when the primary voltage increased, the secondary voltage decreased instead of increasing. Finally, they applied an AC voltage to the da and dn coils on the secondary side of the medium-voltage transformer and measured the primary voltage with an electrostatic voltmeter; both voltages were zero. Based on these tests and data, the personnel initially suspected a short circuit within the electromagnetic unit. However, due to the lack of more detailed technical and testing data for these CVT models, the specific fault could not be determined at that time. Given the need to put the equipment into operation as soon as possible (the combined filter for the station's carrier communication is connected to this CVT), the maintenance personnel removed the CVT. Our bureau's materials company notified the equipment manufacturer's office in Zhengzhou, and a new capacitive voltage transformer was delivered the next day. The new CVT was the same model as the original, except that the structure of the electromagnetic unit was slightly different. With this new CVT as a reference, the staff retested the DC resistance of the coil of the electromagnetic unit of the removed CVT and the voltage applied to the secondary side. The comparative test data is shown in Table 1. These data indicate that a short circuit occurred in the primary winding of the medium-voltage transformer of the electromagnetic unit of the capacitive voltage transformer on the T-line side. Therefore, the staff quickly tested and installed the new CVT, put it into operation in a timely manner, and transported the old CVT back to the bureau for disassembly and inspection. 2 CVT Disassembly Inspection and Fault Cause Analysis In April 2001, our bureau's professional technicians and CVT manufacturer personnel together disassembled and inspected the removed CVT. After the workers loosened several bolts on the flange of the electromagnetic unit's oil tank with a wrench, a pungent and eye-catching oil fumes sprayed out from the flange gaps, and a significant amount of internal pressure could be felt. After removing all the bolts, the capacitor unit was slightly lifted off the lower section of the oil tank using an overhead crane. When removing the lead wire between the intermediate voltage terminal A' and the lower terminal δ of the medium-voltage capacitor C2 and the electromagnetic unit, it was discovered that one of the four bolts securing the lower terminal δ of the medium-voltage capacitor C2 was missing. Because the oil tank was quite full, it was impossible to see where the bolt had fallen. The workers slowly pumped the oil out of the tank. When the oil level was below the terminal block of the medium-voltage transformer, it was finally clear that the fallen bolt had landed among several terminals of the primary winding tap of the medium-voltage transformer. At the point where the bolt contacted the terminal, slight traces of short-circuit welding were found. The oil in the tank had lost its proper pale yellow color and turned a dark brown, like soy sauce. During the pumping process, gas continuously escaped from the oil, and dark brown foam appeared on the surface. After all the oil was drained, it was observed that the iron core of the medium-voltage transformer had been burned away, losing the characteristic luster of the silicon steel sheets. The outermost silicon steel sheets were deformed and bulged in the middle. The white cloth tape covering the windings of the medium-voltage transformer had been burned into black charcoal, and slag fell off when touched. The inner wall of the oil tank was covered with oil stains containing carbonaceous material, leaving a black residue to the touch. To remove the leads of the compensating reactor, the staff removed the cover plate above the terminal box and found that the cover plate had bulged due to excessive internal pressure. At this point, the fault of the CVT was very clear: the primary coil of the medium-voltage transformer had burned out. Even so, we still had the oil maintenance personnel take oil samples for oil chromatography analysis. The analysis results showed that, except for acetylene which was zero, total hydrocarbons and hydrogen levels were significantly higher than the warning values; the calculated three-ratio was 020, indicating a fault type of low-temperature overheating (150-300℃), which further confirmed the fault. Based on the findings of the CVT disassembly inspection, our technical personnel and the equipment manufacturer's personnel unanimously agreed that the cause of the intermediate unit burnout was that a bolt securing the lower end of the medium-voltage capacitor C2 fell into the terminal block of the primary winding of the medium-voltage transformer, short-circuiting part of the primary winding turns. This reduced the AC impedance, causing the primary current to exceed the rated value, resulting in the primary winding burnout. However, the short circuit caused by the bolt was not too severe, or rather, the number of turns short-circuited by the bolt was not large. If the short circuit were severe, the heat generated by the short-circuit current would cause the transformer oil to decompose and release a large amount of gas in a short time, which could potentially cause the downstream tank to explode, or cause the voltage applied across the high-voltage capacitor C1 to become too high, leading to its explosion. As for why this bolt fell off during operation, we believe it was due to improper tightening during installation, which was not detected during inter-process inspections. After the equipment started operating, it was located in the alternating electromagnetic field of the medium-voltage transformer. Under the influence of the alternating electromagnetic field, it continuously vibrated, rotated, and shifted downwards, eventually falling off and causing a short circuit in the primary winding of the intermediate transformer. Fortunately, this problem was discovered during the inspection of the automatic relay protection device, and it was replaced in time, preventing a more serious equipment accident. Source: Power Transmission and Distribution Equipment Network 3 Lessons Learned Capacitive voltage transformers are widely used in power systems, but a failure caused by a loose bolt, as seen in this case, is extremely rare. For power equipment manufacturers, even a one in ten thousand defective products represent a 100% potential hazard for users. This CVT on the T-Jin line was fortunate to be discovered in time, preventing a larger equipment accident. Therefore, as power equipment manufacturers, installation personnel must strengthen their sense of responsibility, and quality inspectors must strictly control acceptance to ensure the quality of every product. On the other hand, the removed part was the electromagnetic unit of that CVT. From the perspective of convenient operation, there are some unreasonable aspects in its structure that need improvement. The lower section oil tank lacks oil level indicators to observe the oil color and level, making it impossible for operators to see the oil color and level; the oil tank also lacks an oil filling hole and oil sampling valve, making it impossible for maintenance personnel to periodically take oil samples for analysis and refilling. The low-voltage terminal XL of the compensating reactor cannot be disconnected from "ground," causing problems for testing. The damper is externally mounted; because the CVT is manufactured with the capacitor divider, electromagnetic unit, and damper adjusted for errors, the dampers are not interchangeable. This causes inconvenience for CVT storage, transportation, installation, and commissioning. Ideally, an internal damper should be used. These issues should be considered by capacitive voltage transformer manufacturers in future product designs. Improvements to the structure of the CVT electromagnetic unit by equipment manufacturers will facilitate equipment operation and maintenance. During inspections, operators can observe the oil status through the oil level indicator, allowing for early detection of problems. Testing personnel can use the oil sampling valve to periodically or after problems are discovered to take oil samples for analysis, enabling early detection of latent faults within the CVT or determination of the nature of the fault. Furthermore, if the BZT device at Jinyuan Substation can be put into operation sooner, and the voltage of the backup line is connected to the device via the instrument transformer, then when the instrument transformer fails and the secondary voltage disappears, the BZT device will report the "backup voltage has disappeared" indicator, and the fault can be detected earlier.