Specifying Generator Testing

International Electric Machines and Drives Conference

Massachusetts Institute of Technology
Cambridge MA, June 2001
J. E. Timperley, Staff Engineer
American Electric Power Pro Sew
Columbus, Ohio

Abstract--Purchasing any generator is a major 11. FACTORY TESTING

investment. What tests the generator is expected to pass
to prove suitability for service and verify quality control Numerous routine tests are conducted on all generator
during manufacturing should be discussed between the components during production and assembly in the factory. It
manufacturer and purchaser and included in contract was common in the first part of the last century to test run an
documents. The following is a list of tests that should be assembled machine in the factory before shipment. This
considered. practice is sometimes followed today with smaller air-cooled
machines but is frequently impractical with larger machine
Index Terms-Generator testing, proof testing, rotor sizes. While a large generator may be fully assembled to
testing, stator testing verify machined surfaces match, running tests are often
deferred to an on site evaluation after a generator is fitted to
I. INTRODUCTION the prime mover.

There was a basic change in production, size and Today some suppliers are reducing production time and
ownership of turbine generators as the 20thcentury closed. expense by eliminating assembled generator testing even on
Previously hydrogen filled machines of 400 MVA and larger the smaller sizes. There are however several factory tests that
were purchased by established public utility companies for can be applied to major generator components such as the
operation at large central coal fired or nuclear power plants. stator and rotor.
Specifications for performance and acceptance testing to
verify a design were often detailed and specific in A few of these in factory tests are:
expectations. Today the majority of generators are driven by
combustion turbines, usually air-cooled and rated 300 MVA Stator (Armature)
or less. The new owners of these ‘backaged power modules”
may be investors planning to capitalize on a deregulated 1. Armature insulation resistance
power market, with limited access to technical expertise. 2. Armature conductor resistance
Testing specifications may be minimized to reduce first costs 3. Armature dielectric proof test
and factory production time. Frequently a supplier is selected 4. Armature mechanical resonance
on technical merits of the turbine; utility or consulting 5. Armature temperature devices
engineers familiar with generators may have limited input
into the drafting of specifications for expected generator
performance and acceptance testing. Details in machine test Rotor (Field)
specificationsmay be left to the supplier.
Today generator suppliers are global in nature, often 1. Winding insulation resistance
utilizing factories in several counties to supply demand. 2. Winding conductor resistance
Each will have an in-house quality control system and these 3. Winding dielectric proof test
are frequently linked to international requirements such as 4. Impedance test for shorted turns
IS0 9001. Competition is extreme with first cost and quick 5. Flux probe test at speed for shorted turns
delivery frequently the major force in manufacturer selection. 6. Mechanical balance verification
This pressure to reduce costs and reduce cycle time has 7. Overspeed
forced manufacturers to restructure production schedules or
face loss of business. Generator testing time in the factory The following is a brief description of these tests and
and in the field is a cost that has been reduced. minimum acceptance criteria. Several of these tests are listed

-1- 139
-0/01/$10@2001IEEE
0-7803-7091
in ANSI C50.13, Table 4 for generators in the factory but not
yet fully assembled.

Ill. STATOR
IV. ROTOR
A . Armature Insulation Resistance
A . WindingInsulation Resistance
Prior to the final winding overpotential test, the insulation
resistance over a ten-minute period is measured with a 1,000- Prior to performing an overpotential test of the field
volt insulation tester. The values are recorded after 30 winding the insulation resistance is to be measured with a
seconds then at one minute intervals for ten minutes. A 500-volt insulation tester over a ten-minute period. Values
minimum acceptable value of resistance after ten minutes is are to be recorded after 30 seconds then every minute up to
100 megohms. Usually this value is several thousand 10 minutes. A minimum acceptable value of insulation
megohms. In addition, a polarization index (the ratio of the resistance is 100 Megohms. In addition, a polarization index
ten minute and one minute readings) is to be calculated. A is to be calculated ( ten minute reading /one minute reading)
minimum acceptable polarization index is 1.5. with a minimum acceptable value of 1.25. If the rotor
operates at 250 volts and higher a 1,000-volt insulation test
B. Armature WindingResistance Measurement should be requested.

The DC resistance of each phase is measured with a low

resistance bridge along with the winding temperature. The B. Winding Conductor Resistance
measured resistance value is converted to an equivalent value
at 25" C. This converted value must be within the After complete assembly the field winding DC resistance
engineering specification of the design value. is measured with a low resistance bridge. An accurate
measurement of the winding temperature is also required.
C. Armature Winding Over Potential Test The measured resistance is then converted to the 25OC value.
There should be no more than 2 % difference from this value
The armature windings are tested in accordance with IEEE and the design value.
standard 115 and ANSI C50.10 by grounding two phases and
applying a power frequency AC voltage whose Rh4S value is C. Winding Dielectric Proof Test
twice rated voltage plus 1000 volts. This is repeated for each
phase for one minute at the test voltage. No dielectric failure An insulation dielectric proof test must be conducted to
of the winding is permitted. insure the winding insulation is suitable for service. The
IEEE Standard 1 15 and ANSI C50.10 request ten times rated
field voltage be applied as an ac voltage for one minute. No
D. Armature Mechanical Resonance failure of the insulation is acceptable.

It is important no part of the stator winding has a

mechanical resonant frequency near or slightly above the D. Impedance Test
magnetic dnving frequency of 100 or 120 Hz. An impact test
to measure the actual resonant frequencies of all endwinding The power frequency (50 Hz or 60 Hz) impedance of the
components should be performed. No resonant frequencies rotor windings should also be measured at several voltages
near or slightly above the driving frequency should exist. up to 120 volts. This should be completed before and after
Resonance above operating frequency are especially the rotor balancing operation to confirm the absence of
dangerous since at operating temperatures this mechanical shorted turns. There will be a measurable drop in winding
resonance will become lower in frequency and could fall into impedance when a turn-to-turn short develops. A running
the critical range. impedance test as the rotor is accelerated to speed and again
as it is decelerated should also be requested. With this test
there will be a step change in impedance as the short
E. R T D 's and Thermocouples develops or clears.

A functional evaluation of all resistance temperature

detectors (RTD's) and thermocouples should be conducted to E. Flux Probe Test at Rated Speed
insure there are no shorted or open circuits and that these
devices are operational and accurate. All temperature sensors After the rotor has been balanced excitation is applied to
should read within a few degrees of each other with a stator the field winding and flux measurements taken to detect tum-
on the factory floor.

140 -2-
to-turn shorts. This should also be repeated after the over 5. ANSI C50.15 Rotating Electric Machinery-Hydrogen-.
speed test to verify no shorted turns have developed. Cooled, Combustion-Gas-Turbine-Driven,Cylindrical-
Rotor Synchronous Generators-Requirements

F. Mechanical Balance Verification These guides describe several tests that should be
requested when purchasing, constructing and installing
Prior to installation a generator field must be balanced at synchronous generators. These are intended to aid the
rated speed. Vibration measurements are taken through the supplier, as well as the owner on what can be expected in the
operating speed range and weight adjustments made until the way of proof tests and assurance of suitability for service.
field exhibits vibration within engineering limits. These standards are periodically reviewed and revised to
Documentation should be provide detailed information on reflect agreements between generator manufacturers and
the critical speeds and the vibration excursions measured at users.
these resonance points. No critical speeds should be
measured near rated operating speed or at the 120% over There are important international standards that can also be
speed. This test should also be requested at both ambient and referenced such as the IEC series. IEC 34-1 for example in
operating temperatures. table 14 lists dielectric test values for both rotor and stator
windings. IEC 34-2 and 34-2A describe several load and no-
load loss evaluation or efficiency tests. IEC 34-4 provides
G. Rotor Over Speed details on methods for determining characteristic
synchronous machine quantities, as transient reactance,
After assembly is complete and the rotor is balanced it is subtransient reactance, negative sequence reactance, and zero
subjected to an over speed as specified in the contract. sequence reactance, from tests. Numerous other tests are also
Normally a rotor is held for one minute at 120% of rated described. Evaluation of the generator enclosure is covered
operation speed. This test is conducted to expose the rotor in IEC 34-5.
components to mechanical forces higher than normal and
demonstrate insulation integrity. Frequently a flux probe test
is conducted to verify that no turn-to-turn shorts developed VI. PARTIAL DISCHARGE MEASURMENTS
during the over speed test. Insulation resistance and
conductor resistance are again performed as well as a There has been significant progress in the application of
winding impedance test to verify that no rotor damage partial discharge measurements to large generators over the
occurred from the over speed. past ten years. There however remains a fundamental
problem of test equipment calibration for accurate
H. Flux Probe Test at Rated Speed measurement of Partial Discharge activity associated with
stator windings. This results in different measurement
After the rotor has been balanced excitation is applied to devices and techniques producing different results on the
the field winding and flux measurements taken to detect turn- same machine. There are several loosely defined
to-turn shorts. This should also be repeated after the over measurement techniques and data analysis methods. No
speed test to verify no shorted tums have developed. No field Partial Discharge based pasdfail criteria has been established
is acceptable for operation if turn shorts are verified to exist. for any of the several voltage classes of generators available.
It is also generally accepted that no single set of
measurements can be interpreted with confidence from a
V. STANDARDS one-time test. Trending of data with the generator operating
under similar conditions is necessary to ascertain the
. There are several standards or guides available on machine condition of a stator winding.
testing that should be referenced when a generator is
purchased. In North America the latest edition of these It is for these reasons that criteria are not yet an industry-
standards should be used: accepted condition as suitability for service with a completed
generator stator.
1. IEEE 1 15, Test Procedures for Synchronous Machines
2. ANSI C 50.10, General Requirements for The evaluation of individual stator bars and coils is
Synchronous Machines however much closer to becoming a useful test to detect
3. ANSI C 50.13, Cylindrical Rotor Synchronous manufacturing defects and evaluate insulation quality. Some
Generators utilities have added Partial Discharge limits as a criterion to
4. ANSI C50.14, Requirements for Combustion Gas their specifications on hydro-generator windings. This
Turbine Driven Cylindrical Rotor Synchronous Generators requirement has proven to be controversial but is gaining
acceptance. Few if any 2-pole or 4-pole stator-winding
suppliers will agree to this acceptance criterion. Testing

-3- 141
continues to refine and reasonable Partial Discharge VI. CONCLUSIONS
acceptance limits will be defined.
The manufacturer should provide along with generator
delivery full documentation of all quality control, suitability
for service and proof tests that were conducted during the
VII. FIELD TESTS manufacturing process. All standards and manufacturing
codes that apply should be referenced. Test data collected
There are several tests that are frequently performed only when a generator in new is valuable as a benchmark for
after a generator has been assembled with the prime mover future maintenance evaluations.
and connected to a power system for loading.
These tests may include: VII. ACKNOWLEDGEMENT

1. Telephone Influence Factor The author gratefully acknowledges the help provided by
2. Efficiency J. R. Michalec during the preparation of this paper.
3. Output Rating
4. Noise Level VIII. BIOGRAPHY
5 . Phase Sequence
6. Vibration Trim Balance James E. Timperley (M 69, SM 84) was born in Omaha
Nebraska, on October 29, 1944. He graduated from
If the generator is of a standard design some of these field Oklahoma State University with a BsEE May 1968.
tests can be omitted and ''typical'' test results can be He began at the American Power
supplied. The expense of commissioningtests other than trim Canton, Ohio Office with the substation engineering grOuP
balance of the rotor and verification of temperature rise at that same year. Activities shifted to establishing procedures
rated load may not be justified. Phase sequence is always at the new Engineering
verified before a generator can be to the power Laboratory in 1970. He then transferred to the Maintenance
system. and Installation group as a rotating machinery specialist in
1973.
Special interests are in the areas of operation, maintenance
and research into diagnostic testing of insulation systems for
motors and generators.

142 -4-