Every plant electrical specification we have seen states a nominal supply of 400 V three-phase at ±[PERCENTAGE] tolerance. Every one of them is also wrong about its own supply, because that tolerance is a regulatory aspiration, not a measured property of the network.
A one-week voltage recording at the low-voltage busbar of a typical industrial site in [LOCATION] will show:
- Sustained excursions to roughly [VOLTAGE] on one or more phases during evening peak, lasting [DURATION].
- Brief excursions to [VOLTAGE] in the early morning as industrial load drops faster than network regulation responds.
- Phase voltage unbalance routinely between [RANGE], against an equipment-terminal limit of 2%.
- Short sags to [VOLTAGE] or below on one phase several times daily during faults elsewhere on the network.
None of this is unusual. All of it is brutal to induction motors and to the availability of any plant whose process depends on them.
The slow failure
A motor fed continuously at [VOLTAGE] instead of 400 V draws roughly [PERCENTAGE] more current for the same shaft torque. Resistive losses rise by roughly [PERCENTAGE]. Winding temperature rises. Insulation thermal life halves for every [TEMPERATURE] of sustained over-temperature.
This is not catastrophic; it is slow. The motor does not trip the day the network sags. It runs hotter, every day, for [DURATION], until the insulation fails at an inconvenient time. The plant logs “motor failed” and replaces it. The plant does not log “the network supplied this motor with abusive conditions for a year, observed on no monitoring system.”
Voltage unbalance is worse. A 4% voltage unbalance produces on the order of [PERCENTAGE] current unbalance, and the resulting negative-sequence rotor heating was never in the motor's design basis. A motor on a persistently unbalanced supply has a thermal life that is a fraction of its nameplate.
The engineering we owe these motors
Tier 1 — design specification
Specify a 1.15 service factor, not 1.0. Specify Class F insulation with Class B temperature rise, giving [TEMPERATURE] of thermal margin. Specify winding thermistors wired to a protection relay that trips on winding temperature, not only overcurrent. This adds [PERCENTAGE] to motor cost and roughly doubles realistic in-service life on this grid.
Tier 2 — motor protection
Replace thermal-overload relays with electronic motor protection monitoring current and voltage unbalance, undervoltage, phase loss, earth fault and starts-per-hour, with logging and thresholds that correspond to motor harm. Cost is [VALUE] per critical motor; the relay repays itself the first time it prevents a rewind.
Tier 3 — voltage support at the site
Where the supply is genuinely outside any motor-tolerable envelope, bring it inside before it reaches the equipment:
- Off-load tap changing on the incoming transformer, set to the measured average, not nominal.
- Automatic voltage regulation on the critical-load bus only, not the whole site.
- On-load tap-changing transformers where the daily swing makes a static tap useless.
- Power-conditioning topologies for genuinely sensitive controllers and instrumentation.
Before any of the above
Capital on voltage support without data is a gamble. The first step is always measurement: a one-week class-A power quality recording at the incomer, and a parallel recording at one or two critical motor terminals to characterise what the load actually sees after the plant's own distribution impedance.
The principle. Motor failures enter the ledger as discrete events; the true cost includes lost production, expedited replacement freight and the second motor halfway through the same slow failure. On the sites we have analysed, the all-in cost of an unprotected motor fleet on this grid reaches [PERCENTAGE] of annual revenue once lost production is properly attributed. We owe our motors a supply they can survive; the instruments to verify it cost less than the third failure.