Is “Lightning Damage” in Variable-Frequency Drives Always Caused by Lightning?

What technicians really mean by “lightning damage”
When maintenance crews find a charred input rectifier or exploded MOV inside a VFD, the report usually says “struck by lightning.” In practice, the phrase has become shorthand for “any transient that exceeded the drive’s surge rating,” not proof that a bolt hit the feeder. Several forensic studies show that more than half of these cases occur on days when regional lightning maps record zero strikes within 10 km of the plant .

Grid-side switching surges
Utility re-closers, power-factor capacitor banks and upstream transformer tap-changers all generate steep-front over-voltages. Because the transient travels through the service transformer, it arrives with the same waveform signature as a distant lightning pulse and is just as capable of blowing the front-end of a VFD . Surge arresters on the pole may clip the peak, but their let-through voltage can still be 2–3 times the drive’s 690 V dc-bus limit.

Motor-side regenerative spikes
During rapid deceleration or an overhauling load the motor acts as a generator. If the braking resistor or chopper fails to absorb the energy, the dc-bus climbs within milliseconds. Modern IGBTs avalanche at ≈1 200 V; once that threshold is crossed, the module ruptures and leaves the same burn pattern technicians associate with a lightning hit .

Ground-potential rise and coupling
Large welding sets or nearby phase-to-ground faults can inject kilo-amperes into the plant ground grid. The resulting potential difference travels up the PE conductor and appears across the VFD’s EMI filter capacitors. The failure mode—blown fuses, split capacitors, shorted C-phase—mirrors classic “lightning” damage but is recorded by the SCADA as an earth-fault event, not weather .

Internal bearing currents – a mini lightning storm
Finally, every PWM VFD produces common-mode voltage that charges the motor shaft. When the potential exceeds the grease’s dielectric strength, a 10–50 A arc erupts through the bearing race. Over time these micro-strikes frost and flute the raceways, a phenomenon engineers colloquially call “bearing lightning” even though no external surge is involved .

How to separate real lightning from imposters

Cross-check the failure timestamp with lightning-location data (LLS, ENTLN, Vaisala).

Inspect surge-counter flags on the main SPD; no increment means the energy came from downstream.

Compare the faulted phase with the direction of incoming power; regenerative spikes always raise the dc-bus equally on all three phases.

Measure ground-grid impedance; if it exceeds 5 Ω, expect ground-potential events.

Look for collateral damage on parallel loads—only lightning or severe utility switching affects multiple devices simultaneously.

Prevention checklist
✔ Install coordinated SPDs: Class I at the main switchboard, Class II at the MCC, Class III at the VFD terminals.
✔ Size the braking resistor for worst-case regenerative energy; verify the chopper reacts within 2 ms.
✔ Keep ground impedance below 2 Ω and isolate shield drains at both ends for high-frequency noise.
✔ Fit motor-side dV/dt filters or sinusoidal chokes when cable length >30 m to curb reflection and bearing currents.
✔ Log dc-bus voltage with the drive’s oscilloscope function; any spike >110 % of nominal deserves root-cause analysis before it becomes “another lightning job.”

Bottom line
Lightning is only one member of a family of fast, high-energy transients that can kill a VFD. Treat every “lightning” failure as a symptom, not a diagnosis, and you will eliminate recurring repairs, reduce insurance claims and keep production lines moving—rain or shine.