Troubleshooting grounding fault of ABB VFD for No. 2 high-pressure gas compressor

Continuing to update. I. Troubleshooting of the failure where Compressor No. 2 does not respond when started 2020.09.24 1. In the afternoon, Compressor No. 2 high-pressure compressor was started according to the standard procedure, but there was no rotation speed or current, and it later indicated a fail to start error. 2. The 9 and 11 terminals for the external emergency stop signal at +A12 were identified; this signal appeared intermittently, with a resistance of 84 ohms when there was a signal and 0 ohms when there was no signal. The source of this signal was sought, and it was determined that there were multiple sources for it, including the emergency stop knob on the VSD cabinet, the on-site LCP emergency stop button, and the shutdown signal from the central DCS system. After checking each of these possibilities, the source of the signal could not be found. 3. At night, terminals 9 and 11 were briefly short-circuited; startup was successful. A no-load test was conducted for two hours with no issues, after which the machine was shut down manually.

2020.09.25 1. In the morning, further inspections were carried out on site, but nothing was found. So, the upstream side of US-C621011 was checked – this is the emergency shutdown signal line that sends signals from the DCS to the VSD. By following this trail, relay SR-02 was identified; it was suspected to be the problem. The connection was switched to SR-03 below it, but there was still no connection. The connection was then switched back to SR-02, and the wire ends were stripped so that the wires could be connected directly to the relay and tightened. On the VSD side, terminals 9 and 11 indicated a proper connection, and attempting to start the device succeeded on the first try. Analysis summary: 1. A small number of issues may remain during the project’s construction phase; these will gradually become apparent during operation, allowing for their identification and resolution.

Thank you to the original poster for sharing the case, but I believe there are some issues with the approach taken to handling this fault. For such high-voltage electrical equipment, we think a more cautious attitude is necessary. The first time an alarm occurred due to grounding, it was reset immediately; the second time, only the cables were checked; and it was only after an explosion occurred inside the cabinet that a thorough inspection was carried out. This approach can easily lead to the escalation of the problem.

Thank you! That makes a lot of sense. At that time, we did react a bit hastily in order to resume production.

II. Repair of grounding issue on the main circuit board of Compressor No. 2 and thyristor breakdown fault 2020.9.26 1. At 8 a.m., Compressor No. 2 stopped operating due to the grounding fault on its main circuit board that had occurred in July: 1stEarthFaultMCirc. A thorough inspection revealed no issues. I tried charging it three times, and there were no more alarms or abnormalities; it seems to be in better condition than it was in July. 2. Inspections were conducted in the morning and afternoon on the insulation of the 36 cables running from the transformer to the VSD, as well as the 12 cables running from the VSD to the motor; no issues were found. 3. In the afternoon, the control motherboard cover was opened to inspect the three circuit boards; meanwhile, a connection was established with the manufacturer’s staff in China, who provided guidance while conducting inspections. The manufacturer believed that the issue could not be determined without charging, so during the fourth charging test, the NP fiber optic connector of the EAF grounding fault detection board was removed. After powering it on for about 10 seconds, the same problems as in July arose: a level 2 grounding fault, a diode failure in phase A, a three-phase short circuit, tripping and power loss at the 11kV terminal of the NGP main transformer, and unusual noises inside the VSD cabinet. Suspend power-on testing. Analysis summary: 1. In the 4th charging test, the phenomena were similar to those in July. It remains to be determined whether the issue lies in the EAF board interface, in a breakdown of some thyristors inside, or in some other new problem; further inspection is required. To be continued

2020.09.27 1. Open the cabinet, remove the control boards for each phase, and check and measure the thyristors in each phase one by one. 2. Upon checking phase A, it was found that the newly installed thyristor gate had been broken down, while the other thyristors and diodes were functioning normally. 3. Inspection of the thyristors and diodes in phases B and C revealed no abnormalities. Analysis summary: 1. It is suspected that there is a problem with the newly installed thyristor itself, which caused a level 1 grounding fault during operation; the specific mechanism requires further consultation and verification. To be continued

2020.09.28 Continued to contact the ABB manufacturer’s engineers. Manufacturer’s opinion: 1. Although the resistance between the gate and cathode of the thyristor in phase A remains high even after compression, it is not damaged and can still be used. 2. The short circuit last time was likely an active protection measure taken by the system due to an excessively high neutral point voltage; the components themselves were not damaged. 3. Both ground fault alarms and short-circuit protection can be related to optical fibers; however, since our system software is unable to generate an alarm by performing a self-check on the rectifier module when only control power is supplied, it is recommended to move the 11kV distribution cabinet to a testing position in order to simulate power supply and determine whether a loose optical fiber connection could cause a ground fault during the self-check. Analysis summary: 1. As recommended by the manufacturer, it is planned to set the VCB of Compressor No. 2 in test mode without actually applying power, in order to conduct self-tests on the fiber optic interface of the ground fault detection module under three conditions: disconnected, connected, and partially connected. This will help determine whether any alarm signals are generated, such as those indicating a ground fault on the main circuit board or diode short circuits. To be continued

2020.09.29 1. In the morning, with compressor No. 2 kept in its test mode, self-check tests were conducted on the NP fiber optic interface of the A-phase EAF grounding fault detection board under three conditions: connected, disconnected, and partially connected. When connected, no alarm was issued; when disconnected, three-phase short-circuit protection triggered due to a level 2 grounding fault; when partially connected, no alarm occurred (as it is difficult to simulate and control the degree of partial connection). 2. Based on the above tests, it can be generally concluded that: 1. No ground fault occurs when powered on ； 2. Disconnection will cause a level 2 grounding fault (more severe than level 1), but it is not yet certain whether an intermittent connection will result in a level 1 grounding fault ； 3. The short-circuit protection trip did not cause damage to the VSD components. 3. Now, secure the fiber optic interface that is questionable using tape to maintain an intact connection. Manufacturer’s opinion: 1. Regarding the situation in Phase C in July, the manufacturer believes that one possibility cannot be ruled out: due to the poor condition of that thyristor, a level 1 grounding fault occurred during operation. When power was supplied again, the condition worsened and the thyristor broke down, resulting in a level 2 grounding fault and activation of the short-circuit protection; normal operation was restored after the component was replaced. Analysis summary: 1. For this secondary grounding fault alarm, it has been confirmed that the fault originated from a disconnection in the optical fiber communication of the EAF grounding fault detection module on phase A; therefore, it is inferred that the primary grounding fault was likely caused by a less severe loose connection. 2. Further research of the relevant data (as shown in the table below) provides the following explanation: 1) Level 1 ground fault: The EAF board detects a NP neutral point grounding voltage that exceeds the limit value, which triggers a signal to the INT board to shut down the unit. This can be explained by the poor condition of the thyristors during operation in July (when a NP voltage higher than the limit was detected), as well as by the loose connection at the NP terminal in this case (a detection signal is present, but it is unstable) ； 2) Level 2 ground fault: A ground condition was detected in the power supply circuit. The possible causes include grounding, faulty circuit boards, and communication issues. Grounding can explain the eventual damage to the thyristor during the charging test in July (which led to grounding), while communication issues can account for the disconnection of the NP port during this test (with no detection signal). 3. However, if there are issues with other components, they need to become apparent during operation. In the near future, consider taking the main transformer out of service and transferring the load in order to supply power to test Compressor No. 2.

2020.09.29 1. In the morning, with compressor No. 2 kept in its test mode, self-check tests were conducted on the NP fiber optic interface of the A-phase EAF grounding fault detection board under three conditions: connected, disconnected, and partially connected. When connected, no alarm was issued; when disconnected, three-phase short-circuit protection triggered due to a level 2 grounding fault; when partially connected, no alarm occurred (as it is difficult to simulate and control the degree of partial connection). 2. Based on the above tests, it can be generally concluded that: 1. No ground fault occurs when powered on ； 2. Disconnection will cause a level 2 grounding fault (more severe than level 1), but it is not yet certain whether an intermittent connection will result in a level 1 grounding fault ； 3. The short-circuit protection trip did not cause damage to the VSD components. 3. Now, secure the fiber optic interface that is questionable using tape to maintain an intact connection. Manufacturer’s opinion: 1. Regarding the situation in Phase C in July, the manufacturer believes that one possibility cannot be ruled out: due to the poor condition of that thyristor, a level 1 grounding fault occurred during operation. When power was supplied again, the condition worsened and the thyristor broke down, resulting in a level 2 grounding fault and activation of the short-circuit protection; normal operation was restored after the component was replaced. Analysis summary: 1. For this secondary grounding fault alarm, it has been confirmed that the fault originated from a disconnection in the optical fiber communication of the EAF grounding fault detection module on phase A; therefore, it is inferred that the primary grounding fault was likely caused by a less severe loose connection. 2. Further research of the relevant data (as shown in the table below) provides the following explanation: 1) Level 1 ground fault: The EAF board detects a NP neutral point grounding voltage that exceeds the limit value, which triggers a signal to the INT board to shut down the unit. This can be explained by the poor condition of the thyristors during operation in July (when a NP voltage higher than the limit was detected), as well as by the loose connection at the NP terminal in this case (a detection signal is present, but it is unstable) ； 2) Level 2 ground fault: A ground condition was detected in the power supply circuit. The possible causes include grounding, faulty circuit boards, and communication issues. Grounding can explain the eventual damage to the thyristor during the charging test in July (which led to grounding), while communication issues can account for the disconnection of the NP port during this test (with no detection signal). 3. However, if there are issues with other components, they need to become apparent during operation. In the near future, consider taking the main transformer out of service and transferring the load in order to supply power to test Compressor No. 2.