Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 61954: Power Electronics — Test Method for Voltage Sag Immunity of Semiconductor Valves
✅ Standard at a Glance
IEC 61954 specifies the test methods for verifying voltage sag immunity of semiconductor power valves used in static var compensators (SVC), STATCOMs, and other power electronic systems connected to AC transmission and distribution networks. Voltage sags caused by remote faults, transformer energization, or large motor starting can impair the firing circuits and control electronics of thyristor valves, potentially leading to commutation failures or valve blocking. This standard defines the critical test parameters, test circuits, and acceptance criteria to ensure valves ride through such disturbances without maloperation. Prepared by IEC Technical Committee 22 (Power electronic systems and equipment).
IEC 61954 specifies the test methods for verifying voltage sag immunity of semiconductor power valves used in static var compensators (SVC), STATCOMs, and other power electronic systems connected to AC transmission and distribution networks. Voltage sags caused by remote faults, transformer energization, or large motor starting can impair the firing circuits and control electronics of thyristor valves, potentially leading to commutation failures or valve blocking. This standard defines the critical test parameters, test circuits, and acceptance criteria to ensure valves ride through such disturbances without maloperation. Prepared by IEC Technical Committee 22 (Power electronic systems and equipment).
🔌 1. The Engineering Challenge of Voltage Sag Immunity
1.1 Why Voltage Sags Matter for Power Electronic Valves
Voltage sags (also called voltage dips) are short-duration reductions in RMS voltage, typically lasting from half a cycle to several seconds. For a thyristor-based SVC valve, the firing pulse generator derives its timing reference from the AC line voltage via a phase-locked loop (PLL). During a voltage sag, the PLL may lose lock, the gate drive power supplies may drop below the thyristor turn-on threshold, and the protective firing circuitry may misinterpret the disturbance as a critical fault. The result can be a complete loss of reactive power support from the SVC at the very moment it is most needed by the grid.
The voltage sag immunity test defined in IEC 61954 is designed to validate that the valve control system can maintain proper firing synchronization and gate drive integrity under defined sag profiles. This is fundamentally different from dielectric withstand tests (which verify insulation) or operational tests (which verify steady-state performance). The sag test targets the auxiliary power systems, control electronics, and firing circuits — the subsystems most vulnerable to voltage disturbances.
| Sag Parameter | Test Range per IEC 61954 | Impact on Valve Operation | Critical Threshold |
|---|---|---|---|
| Retained voltage depth | 10% to 90% of nominal | PLL synchronization, gate drive voltage | <30% retained voltage: high risk of valve blocking |
| Sag duration | 20 ms to 3 s (1 to 150 cycles at 50/60 Hz) | Power supply holdup time, thermal stress on buffers | >100 ms: auxiliary supplies may collapse without energy storage |
| Phase-angle jump | 0° to ±60° | PLL transient response, firing angle correction | >30°: risk of incorrect firing order in 12-pulse converters |
| Point-on-wave initiation | 0° to 360° (any voltage zero crossing) | Transformer inrush current, valve current distribution | Initiation at 90°: worst-case for DC component in transformer flux |
| Multiphase vs. single-phase sags | Three-phase, two-phase, single-phase | Unbalance compensation, negative sequence currents | Single-phase sag: most severe for delta-connected valve groups |
💡 Engineering Insight
The most challenging aspect of voltage sag immunity design is not the magnitude of the sag but the phase-angle jump that accompanies many real-world sags. When a fault occurs on a transmission line, the ratio of line resistance to reactance (R/X) of the fault path changes the phase angle of the retained voltage. Thyristor valves using zero-crossing-based PLLs can experience a transient error of up to 30° to 40° during the sag, which for a 12-pulse converter can cause inter-phase circulating currents. IEC 61954 requires testing with specified phase-angle jumps to ensure the PLL recovery algorithm is robust.
The most challenging aspect of voltage sag immunity design is not the magnitude of the sag but the phase-angle jump that accompanies many real-world sags. When a fault occurs on a transmission line, the ratio of line resistance to reactance (R/X) of the fault path changes the phase angle of the retained voltage. Thyristor valves using zero-crossing-based PLLs can experience a transient error of up to 30° to 40° during the sag, which for a 12-pulse converter can cause inter-phase circulating currents. IEC 61954 requires testing with specified phase-angle jumps to ensure the PLL recovery algorithm is robust.
1.2 Scope and Applicability
IEC 61954 applies to all semiconductor valves used in power electronic equipment connected to AC systems, including SVC thyristor valves, STATCOM voltage-sourced converter (VSC) valves, series compensation thyristor-switched capacitors (TCSC), and high-voltage DC (HVDC) converter valves. The standard covers both line-commutated and self-commutated (forced-commutated) valve types. For each valve type, the critical voltage sag parameters that affect commutation and control differ, and the standard provides guidance for tailoring the test conditions accordingly.
🔌 2. Test Methodology and Circuit Configuration
2.1 Test Circuit Principles
The standard defines two principal test circuit configurations. The first is the synthetic test circuit, where a low-voltage sag generator (typically a transformer tap-changer or solid-state switching network) produces the required sag profile on the valve’s auxiliary power supply and control voltage inputs, while the main power circuit is energized at reduced voltage. This approach is suitable for type testing of complete valve assemblies in the factory. The second configuration is the operational test, where the valve is operating at nominal voltage and a controlled sag is introduced at the valve bus terminals using a series impedance or tap-changing transformer.
The test circuit must be capable of producing all of the sag types defined in IEC 61000-4-11 and IEC 61000-4-34: Type A (three-phase sag), Type B (single-phase sag), Type C (two-phase sag with phase shift), Type D (two-phase sag without phase shift), and Type E (three-phase sag with phase shift). For each type, the retained voltage magnitude, duration, and point-on-wave initiation must be programmable.
2.2 Test Sequence and Acceptance Criteria
The complete test sequence specified in IEC 61954 consists of a matrix of test points covering the sag parameter space. Each combination of sag depth, duration, and phase-angle jump is applied at least three times to demonstrate repeatability. The standard defines three classes of performance:
Class A — Full immunity: The valve continues normal operation throughout the sag with no firing pulse deviation exceeding ±5° and no protective blocking. All auxiliary systems remain operational.
Class B — Limited immunity: The valve may experience temporary firing angle deviation (up to ±15°) but recovers within 3 cycles after sag clearance. Protective firing (bypass pair firing) is permitted but full blocking is not.
Class C — Fall-safe: The valve blocks during the sag but must recover automatically within 5 cycles after voltage restoration without operator intervention and without damage to any semiconductor device.
⚠️ Design Warning
One of the most common failures observed during voltage sag immunity testing is the undervoltage lockout (UVLO) oscillation of gate drive power supplies. When the auxiliary AC voltage sags below the UVLO threshold of the switch-mode power supply feeding the gate drivers, the supply may enter a start-stop oscillation (hiccup mode) as the sag depth fluctuates. Upon voltage recovery, the gate driver may not be ready to fire the thyristor for several hundred milliseconds — sufficient time for a commutation failure in a line-commutated converter. The solution requires gate drive power supplies with extended holdup time (typically >100 ms at 50% retained voltage) and hysteresis in the UVLO circuit to prevent oscillation. IEC 61954 testing exposes this failure mode reliably.
One of the most common failures observed during voltage sag immunity testing is the undervoltage lockout (UVLO) oscillation of gate drive power supplies. When the auxiliary AC voltage sags below the UVLO threshold of the switch-mode power supply feeding the gate drivers, the supply may enter a start-stop oscillation (hiccup mode) as the sag depth fluctuates. Upon voltage recovery, the gate driver may not be ready to fire the thyristor for several hundred milliseconds — sufficient time for a commutation failure in a line-commutated converter. The solution requires gate drive power supplies with extended holdup time (typically >100 ms at 50% retained voltage) and hysteresis in the UVLO circuit to prevent oscillation. IEC 61954 testing exposes this failure mode reliably.
📈 3. Practical Engineering Considerations and Design Mitigations
3.1 Impact on Valve Design Architecture
The voltage sag immunity requirements of IEC 61954 have significant implications for the architecture of power electronic valves. The valve control electronics, including the PLL, firing pulse generator, and valve base electronics (VBE), must be powered from auxiliary supplies with sufficient energy storage to ride through the defined sags. Three common design approaches are used: capacitor-banked DC bus systems providing 100-500 ms holdup time, uninterruptible power supply (UPS) backup for the VBE and control racks, and sag-tolerant switched-mode power supplies with wide input voltage range (typically 80-264 V AC) and high-efficiency operation down to 40% of nominal input.
For the PLL synchronization, modern valve controllers implement adaptive PLLs with feed-forward compensation that can track the phase-angle jump during the sag transient. Some designs use a secondary synchronization reference from a separate voltage transformer on a different bus section, ensuring that at least one reference remains within the operating range during a local sag event.
3.2 Coordination with System-Level Protection
IEC 61954 testing must be coordinated with the system-level protection schemes. The voltage sag immunity characteristics of the valve should be matched to the expected fault clearance times of the AC network. For example, if the transmission system protection clears faults within 60-80 ms (3-4 cycles at 50 Hz), the valve sag immunity should be tested at durations exceeding this clearance time plus a margin (typically 150-200 ms). This ensures that the valve does not block before the system protection has had an opportunity to clear the fault and restore normal voltage.
💡 Engineering Insight
The 2017 edition of IEC 61954 introduced significantly more stringent test requirements than earlier editions, particularly for phase-angle jump testing and multiphase unbalanced sag testing. Valve designs certified to earlier editions may not meet the 2017 requirements. For SVC upgrades or new installations, engineers should verify that the valve manufacturer has performed the complete test matrix specified in the latest edition, not just the minimum test points. Attention should also be paid to the repetitive sag test — a sequence of sags at 1-5 second intervals, simulating the reclosing sequence of transmission line protection. Some valve power supplies may survive a single sag but fail during the second sag due to incomplete recovery of energy storage elements.
The 2017 edition of IEC 61954 introduced significantly more stringent test requirements than earlier editions, particularly for phase-angle jump testing and multiphase unbalanced sag testing. Valve designs certified to earlier editions may not meet the 2017 requirements. For SVC upgrades or new installations, engineers should verify that the valve manufacturer has performed the complete test matrix specified in the latest edition, not just the minimum test points. Attention should also be paid to the repetitive sag test — a sequence of sags at 1-5 second intervals, simulating the reclosing sequence of transmission line protection. Some valve power supplies may survive a single sag but fail during the second sag due to incomplete recovery of energy storage elements.
❔ Frequently Asked Questions
1. How does IEC 61954 differ from IEC 61000-4-11 (voltage dip immunity for low-voltage equipment)?
IEC 61000-4-11 addresses voltage dip immunity for general LV equipment connected to 50/60 Hz AC power supplies (typically IEC 60364 installations). IEC 61954 is specifically for semiconductor power valves used in HV/EHV transmission systems (typically 10 kV to 800 kV). The test levels, duration ranges, and performance criteria in IEC 61954 reflect the higher reliability requirements and more complex control systems of transmission-grade power electronics. IEC 61954 also includes phase-angle jump testing, which is absent from the generic LV immunity standards.
2. What is the significance of the phase-angle jump test in voltage sag immunity?
Phase-angle jumps occur because the fault impedance (R + jX) of the transmission network shifts the phase relationship between voltage and current during a sag. For PLL-based thyristor firing systems, an abrupt phase shift can cause the firing angle to deviate from the commanded value, potentially leading to asymmetric firing of series-connected thyristors or commutation failure. Testing with phase-angle jumps of up to ±60° ensures the valve control system can track and compensate for these transients.
3. Can a valve that passes IEC 61954 type testing still fail in the field during a voltage sag?
Yes, there are several scenarios where field failures can occur despite passing type testing. The most common are: (1) the actual sag profile in the field may have characteristics (multiple notches, harmonic distortion) not covered by the simplified test waveform; (2) aging of electrolytic capacitors in gate drive power supplies reduces holdup time over the valve’s service life; (3) the interaction between multiple valves in the same station during a sag may produce phenomena not observable in single-valve testing (e.g., DC bias accumulation across the converter transformer). Regular maintenance testing of power supply holdup time is recommended.
4. What are the consequences of failing to meet IEC 61954 Class A performance?
If a valve is specified for Class A but achieves only Class B or C, the utility or system operator must implement compensatory measures. These may include: reducing the SVC/STATCOM operating range during weak system conditions, increasing the fault level at the point of connection (e.g., by switching in additional shunt capacitors), or accepting a higher risk of transient instability during critical fault sequences. In some cases, failure to achieve Class A immunity has been the root cause of widespread voltage instability events, as the loss of reactive power support from multiple SVCs during a regional fault can precipitate a voltage collapse.
