Motor Rewinding & Repair

A burnt-out motor is a forensic document. The pattern of the failure tells you what killed it, and a shop that simply rewinds without diagnosing will hand back a motor that fails the same way again. A single-phasingfailure — one supply phase lost — cooks two of the three phase groups symmetrically while sparing the third. A turn-to-turn short burns a localised spot from insulation that finally gave way. Bearing failure shows as a rotor rubbing the stator (a swirl of damage), and overload evenly darkens the whole winding from sustained over-temperature.

The repair therefore starts with the cause: a single-phasing burn points to a protection or supply fault that must be fixed, or the new winding dies too; even overheating points to overloading, poor ventilation or a clogged cooling path. We log the as-found condition, the resistance balance and the failure signature before a single coil comes out — because the rewind is the easy part, and preventing the repeat is the value.

A motor's life is, to a first approximation, the life of its insulation — and insulation ages with heat. Windings are built to a thermal class (B, F or H) defining the maximum temperature the insulation tolerates continuously. The governing rule of thumb is brutal in its simplicity: every 8–10 °C of sustained over-temperature roughly halves the insulation life. A motor run 20 °C hot does not lose a little life — it loses most of it.

A quality rewind therefore matches or upgrades the insulation system (we routinely rewind to Class F or H with modern enamel and resin), uses the correct slot fill, and cures the varnish properly so the winding resists moisture, vibration and the voltage spikes that VFDs throw at it. Using a lower-grade wire or skipping the vacuum-pressure impregnation saves the shop money and costs the owner years of motor life — invisible at handover, expensive later.

The most expensive mistake in motor repair is invisible: a rewind that returns the motor one or two efficiency points lower than new. The classic culprit is the burn-out oven running too hot when stripping the old winding, which damages the inter-laminar insulation of the stator core and increases iron losses. Add a slightly different wire gauge, turns count or coil pitch and you have a motor that runs hotter and draws more power for the same shaft output — forever.

Because motors are such large energy consumers, that small loss dwarfs the repair cost over the motor's remaining life. A 75 kW motor running continuously, dropped from 94% to 92% efficiency, wastes roughly an extra 13,000 kWh a year — at commercial tariffs, far more than the rewind. Good practice (controlled core stripping, a core-loss test before and after, exact-replica winding data) preserves the efficiency. We treat the core-loss test as non-negotiable, because it is the only honest proof the core survived the strip.

There is a defensible line between repairing and replacing, and a trustworthy shop will tell you which side you are on. For larger motors (roughly above 30–40 kW), a quality rewind that preserves efficiency is almost always cheaper over the life than a new motor, and far quicker than a long import lead time. For small, mass-produced motors the economics can flip — the cost of a careful rewind approaches the price of a new, possibly higher-efficiency unit.

The variables are the rewind quality (does it keep the efficiency?), the price and lead time of an equivalent new motor, the running hours (which amplify any efficiency gap), and the criticality of getting the machine back. We put all four in front of the client rather than defaulting to the option that suits the workshop — because the right answer for a continuously-run 90 kW process motor is rarely the right answer for a spare 4 kW fan.

A finished rewind is only as good as the tests it passes, and the meaningful ones go well beyond "it spins." Insulation resistance (IR) and the polarization index (PI) — the ratio of the 10-minute to the 1-minute IR reading — reveal whether the winding is clean and dry or contaminated and damp. A surge comparison test stresses turn-to-turn insulation that a simple megger cannot see, catching the weak coil before it fails in service. A core-loss test confirms the stripping process did not damage the core.

We finish with mechanical truth: bearings replaced as a matter of course, the rotor dynamically balanced, alignment checked, and a vibration baseline taken to ISO 10816 so the customer has a reference for future condition monitoring. A documented test sheet handed over with the motor is the difference between a repair you can trust and one you simply hope worked.

A motor is specified to a rated power, a duty type and a service factor — not just a kW number. The duty type (IEC 60034 S1 continuous through S8 intermittent) describes how the load actually behaves over time; a motor sized for continuous duty is wasteful on a load that runs in short bursts, and one sized for bursts overheats on continuous duty. The service factor is the short-term overload the motor can tolerate without damage — useful headroom, not a licence to run permanently above rating.

As with generators, oversizing is not safety — it is waste. An oversized motor runs at a low load factor where its power factor and efficiency both fall, drawing more reactive current and costing more to run for the same useful work. The right answer is the motor matched to the real shaft load with sensible margin, chosen against the duty it will actually see.

Because a motor often runs thousands of hours a year, its efficiency class (IEC 60034-30-1: IE1 to IE4) dominates its lifetime cost. The purchase price is a small fraction of what a continuously-run motor spends on electricity over its life, so a higher-efficiency motor that costs more upfront is usually far cheaper to own. The difference between an old IE1 and a modern IE3/IE4 motor of the same rating is real money, every hour, for fifteen years.

We weigh the efficiency class against the running hours: for a motor that runs continuously, IE3 or IE4 almost always wins on total cost; for a rarely-used standby motor the case is weaker. The point is to make the decision on lifetime energy, not sticker price — the same discipline we apply to generators and transformers.

A direct-on-line (DOL) start pulls six to eight times full-load current for the first seconds, which stresses the supply (and, on a genset, can dictate the generator size). The starting method is chosen to manage that inrush against the mechanical needs of the load. Star-delta roughly cuts starting current and torque to a third — fine for loads that start unloaded. A soft starter ramps voltage to limit inrush to ~2-3× and reduce mechanical shock. A variable-frequency drive (VFD) ramps frequency and brings starting current near full-load level while giving full control of speed and torque.

The correct choice depends on the load's torque demand at start, how often it starts, and the supply's tolerance for inrush. On borehole, HVAC and conveyor loads we routinely save a customer a generator frame size — or a tripping headache — simply by choosing the right starter rather than defaulting to DOL.

On pumps and fans, a VFD is not just a soft starter — it is an energy goldmine, because of the affinity laws: flow scales with speed, but power scales with the cube of speed. Run a fan at 80% speed and it draws roughly half the power; throttle the same fan with a damper at full speed and you waste the difference as heat. Replacing throttling/recirculation control with VFD speed control routinely cuts pump and fan energy by 30-50%.

VFDs do demand respect, though. Their fast-switching output stresses winding insulation (dV/dt) and can drive circulating bearing currents that pit bearings, so inverter-duty motors, shielded cable and shaft grounding matter on serious installations. They also inject harmonics back into the supply (see our power-quality guidance). Done right, the energy saving dwarfs these costs; done carelessly, they shorten motor life.

Most motor burnouts are preventable with correct protection, and the cheapest insurance on any motor is a properly set protection relay. The essentials: thermal overload (against sustained over-current), phase-failure / single-phasing protection (the classic killer — lose one phase and the remaining windings cook), phase-imbalance and earth-fault protection, and for critical machines, embedded thermistors (PTC) that sense actual winding temperature rather than inferring it from current.

Protection only works if it is coordinated and correctly set — an overload set too high protects nothing, and one set too low nuisance-trips. We size and set protection to the motor and its starter, because the relay that saved the motor is invisible until the day it does. This is also why a rewound motor that fails again so often points back to a protection or supply fault that was never fixed.