Key Takeaways
Why can a shunt reactor fail right at first energization?
What do low-voltage vs. high-voltage tests really reveal?
How do shielding design and manufacturing tolerances impact dielectric strength?
What visible and internal signs indicate dielectric breakdown?
Which design improvements restored reliability in the 345 kV reactor?
Estimated reading time: 5–6 minutes
- Background
As part of the commissioning process for a 345 kV Gas-Insulated Switchgear (GIS) installation, the system was energized for the first time. The GIS comprised:
- 500 MVA power transformers.
- 100 MVA shunt reactor units, each connected to control line voltage and improve voltage stability during light-load conditions.
Shunt reactors are critical for absorbing reactive power and mitigating overvoltage on long EHV transmission lines. Given their continuous connection to the grid under no-load or light-load scenarios, their insulation systems must withstand sustained AC voltages, switching surges, and temporary overvoltages without degradation.
- Event Summary
During initial energization, the differential protection relay for one of the shunt reactors operated, tripping the unit offline within seconds. No external mechanical damage or oil leakage was observed.
- Initial Post-Trip Assessment
Following the trip, standard low-voltage (LV) diagnostic testing was performed:
- Insulation resistance measurement.
- Capacitance and dissipation factor (tan δ).
- Winding resistance measurement.
- Magnetization current test.
All LV tests were within acceptable limits, indicating no gross insulation or winding faults detectable at low voltage.
Given the unexplained nature of the trip and based on WiseGrid’s experts’ recommendation, a series of high-voltage withstand tests was scheduled to evaluate the main insulation integrity under operational stress conditions.
- HV Testing and Failure
The reactor underwent an applied voltage test at 150% of rated phase-to-ground voltage in accordance with IEC 60076-3 standards. During this test:
- The unit failed catastrophically, with a clear drop in insulation resistance and trip of the test set.
- Audible discharge was detected prior to test termination.
- Internal Inspection Findings
An internal examination, carried out in the presence of the OEM engineering team, revealed:
- Black carbonized marks on the inner wall of the A-phase HV turret, consistent with a localized dielectric breakdown event.
- Conductive debris at the base of the A-phase bushing, suggesting partial discharge erosion products or fractured insulation material.
- Electrical tracking along the insulation of the A-phase lead, confirming that the fault path occurred from conductor to grounded metallic structure.
- Root Cause Analysis
The failure was traced to inadequate electrical stress control around the toroid shield assembly in the A-phase bushing lead path.
Key contributing factors:
- Stress distribution inefficiency: The toroid shield was undersized for the actual operating field distribution, resulting in localized areas of high electric field intensity.
- Design and manufacturing tolerances: Small deviations in assembly alignment, material thickness, or component positioning can significantly affect dielectric stress concentration. In this unit, tolerances likely combined in an unfavorable way, reducing dielectric margin.
- Partial discharge inception: Elevated field intensity may have led to partial discharge activity during the initial energization, initiating insulation degradation before the HV test.
- Lack of additional insulating barriers: The original design did not incorporate an auxiliary insulating barrier in the turret area, leaving the bushing base more vulnerable to direct field stress.
These combined conditions meant that dielectric withstand capability could not be reliably guaranteed under both operational and testing overvoltage conditions.
. 
- Corrective Actions Implemented
The OEM implemented a targeted design modification:
- Larger toroid shield: Increased radial depth to improve electric field grading and reduce localized stress.
- Additional insulating barrier: Placed between the lead and grounded turret wall to further reduce peak stress and eliminate potential discharge initiation points.
- Revised assembly procedure: Enhanced dimensional control during installation to minimize positional deviations.
- Validation of the Modified Design
The modified shunt reactor underwent the full set of Factory Acceptance Tests (FAT) and Site Acceptance Tests (SAT):
- Applied Voltage Test: 150% rated voltage – Pass
- Lightning Impulse Withstand Test – Pass
- Partial Discharge Test (at rated voltage and 1.1 p.u.) – Pass, with PD levels well below IEC limits.
No abnormal discharge activity or insulation deterioration was observed, confirming that the redesigned stress control arrangement achieved the intended dielectric performance.
- Conclusion
The root cause of the 345 kV shunt reactor failure was insufficient electrical stress control at the A-phase toroid shield, compounded by component positional tolerances. This resulted in dielectric breakdown between the A-phase lead and grounded turret during high-voltage stress conditions.
The implemented design modifications—larger toroid shield, additional insulation, and improved assembly precision—have restored dielectric reliability, as confirmed by post-modification HV testing.
