A stator rewind strips the failed windings out of an alternator’s stator core and rebuilds them with new copper and insulation on the original core and frame. When the core and mechanicals are sound and only the insulation has failed, a rewind restores the machine at a fraction of replacement cost. The hard part is proving which case you have — before anyone quotes.
TL;DR
- A rewind replaces the stator’s copper and insulation; the core and frame stay — so the core must be proven sound first
- Insulation degrades under electrical, mechanical and thermal stress plus contamination (Stamford AGN 015) — and, on old machines, plain age
- Meggering separates the recoverable cases — damp or dirty windings — from insulation that is actually finished
- Demand a failure investigation before any quotation; a replacement quoted on sight is the anti-pattern
- The reference job: a 1250 kVA MTU alternator rewound for Siemens at Verna, Goa — investigated first, rewound once, recommissioned on load
What a stator rewind is — and what it is not
The stator winding is where an alternator earns its rating: insulated copper coils, seated in the slots of a laminated steel core, generating the output power. A full rewind strips those coils out, inspects the core, and rebuilds the winding to the original design data — new copper, new slot insulation, new impregnation, baked and tested before reassembly.
A rewind is not a cure-all. Bearings, rotating diodes, AVRs and end-shields are overhaul items, not rewind items — many alternator “failures” never reach the stator copper. And a rewind cannot fix a damaged core: new copper in a core with shorted laminations fails the same way the old winding did. The decision starts with evidence, not with a price.
How stator insulation fails
Stamford’s guidance note on testing winding insulation, AGN 015, names the enemies plainly: electrical, mechanical and thermal stresses, plus environmental contamination, degrade the insulation system throughout its life. The everyday attackers are moisture and surface contamination on the winding overhang. Underneath them all runs age, quietly consuming the margin the insulation was built with.
Each failure mode points to a different remedy:
| What happened | What testing shows | Usual verdict |
|---|---|---|
| Damp winding after long idle | Low IR that recovers after a proper dry-out | Dry out, fix the moisture source — no rewind |
| Contaminated winding | Low IR, both PI readings low, visibly dirty | Strip, pressure-wash, bake, over-coat with a compatible resin |
| Age-weakened insulation | IR stays low after drying; poor polarization index | Rewind — if the core is sound |
| Winding fault — shorted turns, earth fault | Fault located on test, often with visible damage | Rewind or replace, decided by core condition |
| Core damage | Core-loss increase on a loop/core test | Repair laminations where practical; otherwise replace |
The diagnostic pattern in the first three rows follows Stamford’s published guidance; the rewind verdict is ours, proven at Goa. Core testing belongs to the ANSI/EASA repair standard, covered below. None of the rows can be told apart from the outside of the machine.
Meggering: the evidence that decides it
An insulation resistance test — meggering — applies a test voltage of the order of 500 V, measures the leakage current, and reports insulation resistance in megohms. Two disciplines make the number honest.
Disconnect before you test. The AVR senses — and on self-excited machines feeds from — the stator terminals, and a megger’s test voltage can damage it along with any connected instrumentation. Stamford’s AGN 015 devotes pages to which leads come off first.
One reading is not a verdict. Hot windings megger far lower than cold ones — normal, not a finding. What matters is where the value sits on the ladder, and how it responds to drying:
| IR reading | What it typically means (per Stamford AGN 015/040) |
|---|---|
| Above 100 MΩ | Factory-fresh winding |
| Around 25 MΩ | Well-stored, unused machine |
| A few MΩ | Normal for a machine in service |
| Around 1 MΩ | The working floor — reliable service is known down to 1 MΩ, the minimum some marine classification societies stipulate |
At or below the floor, dry the machine out — Stamford’s methods run from a short unexcited run to a controlled short-circuit heat run — and megger again. Recovery means moisture: fit the anti-condensation heater and fix the enclosure’s condensation. A visibly dirty winding with both polarization index readings low means wash-and-bake. The PI test is the megger read at one minute and again at ten; the ten-to-one-minute ratio comes out around 3 or better on a healthy winding.
But if a clean, properly dried winding still meggers low and holds a poor PI, the insulation itself is finished. On a machine decades into its life, the cause is age — and that is precisely what a rewind exists to fix.
The anti-pattern: a quote before a cause
A big alternator fails, a vendor visits, and a replacement — or a rewind — is quoted on sight, before anyone has meggered anything or written down a cause. A repair without a root cause is a repeat failure on a timer. A replacement without one can be worse: if condensation or contamination killed the old machine, the new one inherits the same enclosure and the same fate, at many times the cost of fixing the actual problem.
The protection costs one sentence in your enquiry: findings in writing before scope, scope in writing before price. Any competent repairer can produce megger readings, a PI ratio and a stated cause; one who resists has told you something useful.
Rewind or replace: how to decide
With the investigation done, four questions decide it — none of them answerable before the tests:
- Is the core sound? Decisive. A healthy core makes a rewind viable; a damaged one changes the economics entirely, because core repair is a job on top of a rewind.
- What state is the rest of the set in? A rewind makes most sense on a machine whose engine and mechanicals have life left. If the whole generating set is at the end of its life, money spent on the alternator is better put toward replacement.
- What does availability look like? A large-frame alternator is not a shelf item — a replacement is sourced against manufacturer lead time, and mating a new alternator to an existing engine, baseframe, coupling and controls is its own engineering exercise. A rewind returns the machine you already integrated.
- What is the cost relationship? A rewind on a sound machine comes in at a fraction of replacement cost. There is no honest generic rupee figure — winding data, frame size and logistics set it — which is why the price follows the inspection, never the phone call.
Executing a rewind on a big machine
On big frames, a careless rewind destroys the asset it was meant to save. The risks are known, and so are the controls — the ANSI/EASA AR100 repair standard prescribes most of them:
- Getting the old winding out without cooking the core. Winding removal uses a temperature-controlled burnout with the part’s temperature monitored throughout. Overheat the core and its interlaminar insulation is gone — invisibly.
- Proving the core before and after. AR100 calls for a core (loop) test before and after winding removal, investigation of any increase in core losses, and repair or replacement of damaged laminations — so a bad core is found before the rewind, not after recommissioning.
- Duplicating the winding data. The rewind must reproduce the original electrical design: conductor cross-section held or increased, coil extensions no longer than original, insulation system equal to or better than what came out — Class H or F materials matching the machine’s build.
- Impregnation quality. Large Stamford frames leave the factory vacuum-pressure impregnated with epoxy resin — the VPI autoclave evacuates the slots, floods them under vacuum, then applies pressure to fill every void before a controlled bake. A rewind’s impregnation and cure must be worthy of that; a poorly filled slot is a vibration and moisture problem on a delay.
- Testing without wounding. IR and PI verify the new winding. Hi-pot testing is used sparingly: per IEC 60034-1, as Stamford summarises it, a winding is never again tested at the full factory level — repeats run at no more than 80% of it.
- Recommissioning is part of the job. The machine goes back on the engine, gets its AVR and protections back, and runs on load. The gap in the Indian market sits exactly here: rewinding workshops that cannot commission a generating set, and rental companies without a rewind workshop.
The rewind we point to: 1250 kVA at Verna, Goa
This decision guide is written from a job we did, not a job we read about. A 1250 kVA MTU set’s alternator failed at a Siemens facility at Verna, Goa. The instinct is to quote a replacement; we ran the failure investigation first. Meggering established that the insulation had been weakened by age, and the root cause went to the customer in writing before any scope was priced. On that evidence, the machine took a full alternator overhaul and stator rewind, followed by recommissioning and load verification — at a fraction of what a replacement would have cost.
The same workshop discipline runs across brands: alternator overhauls on the Perkins sets protecting HSBC Vishakhapatnam — one 1010 kVA and two 1500 kVA machines — and rewind and overhaul work on Stamford, Crompton, Leroy Somer and Kirloskar alternators, Class H/F machines up to 1250 kVA and beyond.
Where this sits in your maintenance calendar
The cheapest stator rewind is the one you see coming. Insulation resistance is a trendable number: measured at every C check — where the maintenance ladder says the alternator gets real attention — a slow decline over years announces ageing insulation long before a breakdown does. Stamford’s prevention list is short: keep the winding clean and dry, fit an anti-condensation heater that switches on whenever the set rests, and run the set on load regularly enough to keep moisture driven out.
If your alternator has already failed — or your megger readings are drifting somewhere they shouldn’t — start with the investigation, not the quotation. Send the machine details and any readings you have on WhatsApp, and the repair and overhauls team will tell you what the evidence supports: a dry-out, a wash-and-bake, a rewind, or an honest recommendation to replace.
Sources: Stamford / Cummins Generator Technologies, AGN 015 “Testing Winding Insulation Systems” — https://www.stamford-avk.com/sites/stamfordavk/files/2024-04/AGN015_D.pdf — degradation stresses, megger voltage and AVR disconnection, IR values, dry-out methods, PI method and interpretation, and the IEC 60034-1 repeat hi-pot limit. Stamford AGN 040 “Winding Insulation System” — https://www.stamford-avk.com/sites/stamfordavk/files/AGNs/AGN040_D.pdf — VPI, dip-and-bake and trickle impregnation, epoxy VPI on large frames, IR maintenance measures and wash-bake-overcoat recovery. ANSI/EASA AR100 “Recommended Practice for the Repair of Rotating Electrical Apparatus”, as summarised by Plant Services — https://www.plantservices.com/equipment/industrial-motors/article/11289987/whats-new-in-the-ansi-easa-ar100-2020-motor-repair-standard — temperature-controlled winding removal, before/after core testing and winding-duplication requirements. Siemens Verna Goa and HSBC Vishakhapatnam details are from our published project and service records. Your alternator’s O&M manual governs test voltages and acceptance values for your machine.