Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Surviving the Strike — IEC 60722 Guide to Lightning Impulse Testing
When a power transformer leaves the factory, it carries a nameplate that includes its Basic Insulation Level (BIL) — a number like “350 kV BIL” that represents the lightning impulse voltage it can withstand. That number is not a theoretical calculation. It was verified in a high-voltage laboratory where massive impulse generators discharged millions of volts across the transformer’s insulation in a few microseconds. IEC 60722 (1982) is the engineering guide that defines how those tests should be designed, executed, and interpreted. It bridges the gap between the test requirement (“apply a 1.2/50 μs impulse of X kV”) and the practical reality of generating and measuring that waveform with acceptable accuracy.
Core insight: IEC 60722 is not the test requirement standard (that is IEC 60076-3 for transformers, IEC 60060-1 for general high-voltage test techniques). Rather, it is the guide to test technique — the “how to actually do it” document that explains impulse generator circuit design, voltage divider selection, measurement system calibration, and waveform analysis that the test engineer needs at the bench.
The 1.2/50 Waveform — Why Those Numbers Matter
The standard lightning impulse is defined by IEC 60060-1 as a 1.2/50 μs waveform — a voltage that rises to its peak in 1.2 μs (front time) and decays to half its peak value in 50 μs (tail time). IEC 60722 provides the engineering underpinnings of how to generate and verify this waveform:
| Waveform Parameter | Definition per IEC 60722 | Tolerance | What It Tests |
|---|---|---|---|
| Front time (T1) | 1.67 (t90 – t30), where t90 and t30 are the times at 90% and 30% of peak voltage on the rising edge | ±30% (1.2 μs ±0.36 μs) | Winding series capacitance and turn-to-turn insulation — rapid voltage rise stresses the first few turns disproportionately |
| Tail time (T2) | Time from virtual origin (intersection of the 30%-90% line with the time axis) to the point on the tail where voltage falls to 50% of peak | ±20% (50 μs ±10 μs) | Major insulation (winding-to-winding, winding-to-ground) — the sustained voltage stresses bulk insulation |
| Peak value (Up) | Maximum absolute voltage reached during the impulse | ±3% | Overall insulation integrity at the rated BIL |
| Overshoot | Voltage exceeding Up at the very beginning of the impulse (typically due to stray inductance ringing) | ≤5% of Up (for most applications) | Can cause false failures — overshoot stresses the insulation beyond the intended test level |
Practical reality: Achieving the specified waveform at the transformer terminals is not trivial. A transformer winding is a complex R-L-C network, not a simple resistive load. When the impulse generator (typically a Marx generator) discharges into the winding, the voltage waveform at the transformer bushing can be significantly distorted by the winding’s impedance characteristic — particularly for low-impedance, high-power transformers. IEC 60722 provides guidance on how to adjust the generator’s series and tail resistors to compensate for the DUT’s impedance and achieve the required waveform within tolerance. This is as much art as science, and the standard’s practical guidance is invaluable to test engineers.
Measurement Chain and Calibration
The measurement of a multi-megavolt impulse with microsecond precision requires a sophisticated measurement chain. IEC 60722 provides detailed guidance on each element:
- Voltage dividers: The high voltage must be scaled down by a factor of 10,000 or more before it can be digitized. IEC 60722 discusses damped capacitive dividers (preferred for fast impulses due to superior high-frequency response), resistive dividers (simpler but susceptible to stray capacitance effects), and mixed dividers. Crucially, the standard addresses the proximity effect — the divider’s presence alters the electric field in its vicinity, and if mounted too close to the test object or grounded structures, the measured ratio can deviate significantly from the calibrated ratio.
- Digital recorders and oscilloscopes: The standard specifies the minimum bandwidth (typically ≥60 MHz for 1.2/50 μs impulses), sampling rate (at least 8 samples per μs, i.e., 8 MS/s minimum), and vertical resolution (≥8 bits) for capturing impulse waveforms. These specifications, written in 1982 for analog oscilloscopes with Polaroid cameras, have been updated in subsequent IEC 61083 standards for digital recorders, but the core principles remain: you need enough bandwidth to capture the rapid front, and enough record length to capture the tail.
- Calibration: The entire measurement chain — divider, cable, attenuator, digitizer — must be calibrated as a single system, not component by component. IEC 60722 emphasizes that calibrating the divider alone on a low-voltage bridge and then separately characterizing the digitizer’s step response will miss the interaction effects that only appear when the complete chain is subjected to a full-voltage impulse of known shape.
Engineering insight: The most common root cause of a failed lightning impulse test is not a genuine insulation defect — it is external flashover at the bushing or in the test setup. When the test voltage approaches the BIL level, the clearances between the bushing terminals and nearby grounded structures (walls, gantry, measurement equipment) become critical. IEC 60722 advises minimum clearances (typically 1.3 meters per 100 kV in air at standard atmospheric conditions) and recommends verifying the test setup with a dummy load before connecting the expensive transformer. A failed test due to flashover in the test circuit is an expensive and embarrassing mistake that proper setup verification prevents.
Frequently Asked Questions
- Q1: What is the difference between a “full lightning impulse” and a “chopped lightning impulse”?
- A full lightning impulse (1.2/50 μs) is applied and allowed to decay naturally. A chopped impulse is suddenly interrupted (typically after 2-6 μs) by a flashover across a rod gap or a triggered spark gap, simulating the effect of a lightning arrester operating. The chop creates an extremely steep voltage collapse (easily several thousand kV/μs) that tests the inter-turn insulation far more severely than a full impulse, as the voltage distributes very non-uniformly across the winding during the chop.
- Q2: Why is the Marx generator circuit still the standard for impulse generation?
- The Marx generator (capacitors charged in parallel through resistors, then discharged in series through spark gaps arranged as switches) remains dominant because it is the only practical way to generate megavolt-level impulses with microsecond risetimes. Modern semiconductor switches (IGBTs, thyristors) cannot approach the voltage ratings of a spark gap. A single Marx stage might charge to 100-200 kV, and 10-20 stages in series produce the 1-3 MV needed for testing transmission-class transformers.
- Q3: How does atmospheric pressure and humidity affect impulse test results?
- Significantly. The flashover voltage of air gaps decreases with lower air pressure (higher altitude) and varies non-monotonically with absolute humidity. IEC 60722 references the atmospheric correction factors defined in IEC 60060-1: tests at high altitude (e.g., 1500 m) require a correction factor that can increase the required test voltage by 10-15% relative to sea-level conditions. Laboratories must record atmospheric conditions during every test and apply the correction factor — a step that is sometimes overlooked and can lead to unjustified test failures or, worse, false passes.
