Post-Repair Generator Testing: Standards, Procedures, and Acceptance Criteria

Post-repair generator testing is a critical quality assurance process that verifies the dielectric, mechanical, and thermal integrity of generators after overhauls, rewinds, or component replacements. By applying internationally recognized standards from IEC, ISO, IEEE, and GB, operators can confirm that repaired generators meet acceptance criteria before returning to service. This guide covers sample types, test items, methods, acceptance limits, and practical considerations for generator post-maintenance verification.

Technical Overview

When a generator undergoes repair—whether a stator rewind, rotor field rewinding, bearing replacement, re-wedging, or core re-stacking—its electrical and mechanical characteristics change. Post-repair testing (also called post-overhaul inspection or commissioning verification) provides objective evidence that the repaired machine has restored its dielectric strength, winding integrity, balance, alignment, and thermal performance before reconnection to the prime mover and grid.

The testing philosophy combines static electrical tests, performed while the machine is at standstill, with dynamic tests performed during a first run or heat run. Static tests confirm insulation condition and winding continuity; dynamic tests confirm vibration behavior, temperature rise, output characteristics, and overall operational readiness.

Post-repair acceptance programs are typically built around the following reference frameworks:

• IEC 60034-1 — Rotating electrical machines: rating and performance, including routine test voltages
• IEC 60034-27 series — Partial discharge measurements and dielectric dissipation factor on stator winding insulation
• IEEE 43 — Recommended practice for testing insulation resistance of rotating machinery
• IEEE 95 and IEEE 522 — DC high-potential and impulse (surge) testing practices
• ISO 20816 / ISO 10816 series and ISO 7919 — Vibration evaluation on non-rotating parts and shafts
• ISO 21940 (formerly ISO 1940) — Balance quality requirements for rigid rotors
• GB/T 7064 and GB/T 8564 — Chinese national standards for turbo generators and their erection and acceptance
• IEEE 1068 — Recommended practice for the repair and rewinding of AC electric motors and generators

Because repaired windings may not be retested at full factory test voltage, most specifications and insurer requirements mandate that routine dielectric tests after repair be performed at a reduced level—commonly 75% to 80% of the original factory test voltage—unless a full rewind to new-build specification has been completed.

Testing Samples

Stator Windings and Stator Cores

Stator samples include fully rewound or partially repaired main windings, high-voltage bushings, line-end and neutral-end circuits, and the stator core lamination stack. Testing verifies ground-wall insulation condition, interturn insulation, wedge tightness, and core interlaminar insulation after any re-wedging or core repair work.

Rotor and Field Windings

For synchronous machines, field winding samples include the complete rotor circuit, pole windings, damper windings, slip rings, brush gear, and shaft-mounted rectifiers (for brushless excitation). For induction-type and doubly-fed generators, rotor cage or wound-rotor circuits are evaluated after rewinding or bar replacement.

Complete Assembled Generator Units

Full-machine samples cover turbo generators, hydro generators, wind turbine generators (both doubly-fed induction and permanent magnet synchronous types), and industrial or marine gensets that have completed overhaul in a workshop or on site. These units undergo integrated electrical, mechanical, and thermal verification.

Auxiliary and Replacement Components

Ancillary samples include rolling-element or sleeve bearings, bearing insulation layers, embedded resistance temperature detectors (RTDs) and thermocouples, space heaters, cooling systems, excitation and voltage regulator modules, and terminal box components. Each auxiliary item is checked independently before being energized together with the main machine.

Testing Items

A comprehensive post-repair test plan typically includes the following items, selected according to machine voltage class, rating, and the scope of the repair:

• Insulation resistance (IR) of stator, rotor, and accessory circuits
• Polarization index (PI) and dielectric absorption ratio (DAR)
• DC winding resistance of all phases and field circuits, with phase-to-phase balance comparison
• AC or DC high-potential (hipot) dielectric withstand tests at reduced post-repair voltage
• Surge comparison (impulse) testing for interturn insulation in form-wound and random-wound stators
• Partial discharge (PD) magnitude and pattern measurement on high-voltage stator windings
• Dielectric dissipation factor (tan delta) and tip-up evaluation on epoxy-mica insulation systems
• Open-circuit (no-load) saturation characteristic and residual voltage balance
• Short-circuit characteristic and locked-rotor or impedance checks where applicable
• Rated load or simulated load test with full temperature-rise measurement by resistance method
• Vibration severity on bearing housings (RMS velocity) and shaft relative vibration (peak-to-peak displacement)
• Residual unbalance verification and influence of trim balancing weights
• Air gap measurement and concentricity verification between rotor and stator
• Bearing insulation resistance and shaft voltage / shaft current checks
• Verification of embedded RTD and thermocouple circuits, including insulation to ground
• Phase sequence, rotation direction, and terminal marking verification
• Efficiency, power factor, and voltage regulation confirmation against nameplate data

Testing Methods

Insulation Resistance and Polarization Index

Insulation resistance is measured with a temperature-compensated megohmmeter applying 500 V DC up to 5 kV DC, selected according to winding rated voltage as guided by IEEE 43. Readings are taken at 30 seconds and 60 seconds for DAR, and at 1 minute and 10 minutes for PI. As common acceptance thresholds, PI should be at least 2.0, and minimum IR values are commonly 100 MΩ for modern form-wound stators and (rated kV + 1) MΩ for older insulation systems, always corrected to 40 °C.

Winding Resistance Measurement

DC resistance of each phase or section is measured with a four-wire (Kelvin) micro-ohmmeter after the machine has stabilized at ambient temperature. Results are temperature-corrected and compared with factory or pre-repair records; unbalance between phases should generally remain within 2%, with any step change investigated for joint, braze, or connection defects introduced during repair.

Dielectric Withstand Testing

High-potential tests apply a controlled overvoltage between windings and frame for 60 seconds, following test voltage formulas in IEC 60034-1 or NEMA MG 1. After in-service repairs, the test voltage is typically limited to 80% of the original routine value unless the machine was fully rewound. DC hipot methods per IEEE 95 may be used where AC test capacity is unavailable, with the equivalent DC voltage usually taken as 1.7 times the AC test voltage.

Surge Comparison Testing

A repetitive impulse voltage is injected into each phase while the resultant waveform is compared against a reference phase or a stored pre-repair waveform. Divergence in amplitude, frequency, or damping indicates interturn, coil-to-coil, or phase-to-phase weaknesses that a DC insulation test cannot detect. Impulse test levels for form-wound stators are commonly referenced to IEC 60034-15 and IEC 60034-18 practice.

Partial Discharge and Tan Delta Diagnostics

For machines rated at approximately 3.3 kV and above, offline PD measurement per IEC 60034-27 and IEC 60270 quantifies discharge magnitude (in pC or nC), phase-resolved patterns, and inception/extinction voltages. Tan delta measurement per IEC 60034-27-3, performed in steps up to rated phase-to-ground voltage, evaluates global insulation condition, with the tip-up between specified voltage steps compared against baseline data.

No-Load, Load, and Temperature Rise Tests

The machine is run uncoupled to record the open-circuit saturation curve and confirm balanced residual voltages between phases (typically within 1% to 2%). A subsequent load or heat run at rated output establishes steady-state winding temperature rise using the resistance method in accordance with IEC 60034-1, verifying that repaired windings and cooling paths stay within their insulation class limits.

Vibration, Balance, and Mechanical Checks

Bearing housing vibration is measured as RMS velocity in mm/s and evaluated against ISO 20816 / ISO 10816 zone limits, while shaft relative vibration is assessed per ISO 7919. Air gap is measured at multiple rotor positions to confirm concentricity, typically within ±10% of the average gap. Residual unbalance is verified against ISO 21940-11, with G2.5 being a common balance grade for generator rotors. Bearing insulation and shaft voltage measurements confirm that bearing currents are properly controlled.

Applications

Post-repair generator testing is applied across every sector that depends on rotating machine reliability:

• Thermal, combined-cycle, and nuclear power plants verifying turbo generator overhauls, rotor rewinds, and stator re-wedging projects
• Hydropower stations commissioning hydro generators after pole rewinding, damper bar replacement, or air gap correction
• Wind farm operators validating doubly-fed induction generators (DFIG) and permanent magnet synchronous generators after gearbox-aligned bearing replacement or rewind
• Petrochemical, steel, cement, and mining facilities restarting critical standby and continuous-duty generators
• Data centers, hospitals, and telecom sites performing acceptance testing of repaired emergency diesel gensets before blackout-readiness certification
• Marine and offshore installations complying with classification-society-driven overhaul verification for main and auxiliary generators
• Independent service centers and accredited third-party laboratories providing objective acceptance reports for owners, insurers, and warranty claims

FAQ

Why is testing mandatory after a generator repair?

Repairs alter insulation systems, connections, balance, and alignment. Without verification tests, latent defects such as weak interturn insulation, loose wedges, or contaminated windings could cause immediate failure on restart, potentially destroying a far more expensive machine than the repair itself.

What is a passing insulation resistance value?

There is no single universal number, but IEEE 43 provides widely used minimums—commonly 100 MΩ corrected to 40 °C for modern form-wound stators and 5 MΩ for random-wound low-voltage machines—with the trend between pre-repair and post-repair readings mattering as much as the absolute value.

Why is the hipot test voltage reduced after repair?

Aged insulation that has been disturbed during repair is more vulnerable to overvoltage damage. Industry practice therefore retests repaired windings at approximately 75% to 80% of the original routine test voltage, while fully rewound stators built to new specifications may be tested at full routine levels.

Can surge testing damage an already weakened winding?

Surge testing applies controlled, low-energy impulses within specified voltage limits. When performed at correct levels by trained personnel following IEC 60034-15 guidance, the diagnostic value of detecting interturn faults far outweighs the risk, though heavily degraded vintage insulation may be tested conservatively.

How long does a complete post-repair test program take?

Static electrical tests on a medium-sized industrial generator typically require one to two days, while a full program including no-load run, load heat run, and vibration mapping may extend from several days to two weeks for large turbine generators, depending on cooling and stabilization times.

Is partial discharge testing necessary on every repaired generator?

PD and tan delta diagnostics are strongly recommended for machines rated 3.3 kV and above, or whenever an epoxy-mica insulation system has been partially replaced. They establish a baseline fingerprint that enables trending during future offline outages and help confirm that repair workmanship did not introduce discharge sources.

Who should perform and witness the tests?

Testing should be executed by qualified technicians using calibrated instruments, and for high-value machines it is common practice to have an accredited third-party laboratory or the owner's engineer witness the program, review acceptance criteria, and issue an independent report for warranty and insurance records.