AGN 015 - Testing Winding Insulation Systems
AGN 015 - Testing Winding Insulation Systems
AGN 015 - Testing Winding Insulation Systems
INTRODUCTION
Background
The IR value provides an indication of the condition of the entire electrical insulation system
including the impregnation material. The chemical design and formulation of electrical
impregnation material requires considerations be given to the following aspects:
When the ideal impregnation resin has been formulated, the above requirements are all
satisfied. This means the Chemist has cleverly ensured that all the required mechanical and
electrical characteristics have been achieved and surpassed by ‘safety factor’ multiples of the
fundamental required levels for each characteristic.
Note: The full test procedure is detailed in the Installation, Service and Maintenance Manuals
(also known as the Owner’s Manual) issued with every alternator and can also be found on the
website, www.stamford-avk.com
The IR test involves an instrument commonly called a Megger. This instrument will apply a
‘test voltage’, measure the ‘leakage current’ and then indicate the value of the winding’s IR in
[M] Ohms.
A Megger will subject a voltage of some 500V (or more with some megger instruments) to any
circuit to which it is connected. If a winding is in a very contaminated condition and then
subjected to the typical 500V Megger test the resulting leakage current could promote
permanent damage to the winding insulation system. For this reason, care should always be
taken to observe the readings indicated at the very moment the megger test begins and if
possible, always slowly increase the applied ‘test’ voltage.
The Megger test voltage could damage electrical / electronic components that are connected
to the winding under test and so must be disconnected. The AVR is directly connected to the
stator windings to provide the required ‘Voltage sense’ information. With the SX (Self excited)
type AVR, the excitation power supply is also taken directly from the stator terminals.
Various wiring loom assemblies connect the stator to the AVR, some incorporate individual
‘Voltage sense’ and ‘power’ leads, others interconnect the various AVR inputs by links along
the AVR terminals.
Care must be taken to ensure the loom arrangement is understood and that the easiest
possible point is chosen for AVR disconnection, so ensuring the megger test will not damage
the AVR and that mistakes are not easily made when reconnecting the AVR after the test.
Within the terminal box mounted there may well be optional equipment connected to the stator,
e.g. Radio Interference Suppression kit. These items must also be disconnected.
Every alternator leaves the Cummins Generator Technologies (CGT) factory with an
Installation, Service & Maintenance Manual (Owner’s Manual)> Also enclosed in the package
supplied with the alternator, will be a full set of electrical connection diagrams specific to the
alternator, showing AVR and Stator Connections and also, diagrams for any fitted optional
The following table offers basic guidance regarding AVR connections, but always confirm the
individual alternator connection details before disconnecting any leads and so risk, loosing
individual lead identity.
AVR Type If Sensing T/ F Sensing Lead No’s AVR Sensing AVR Power
is Used : At Stator At AVR Term. No’s. Term No’s.
SX 460 NO 7, 8 7, 8 7, 8 7, 8
SX 440 NO 7, 8 7, 8 3, 2 P2, P3, [P4]
SA 465 NO 7, 8 7, 8 3, 2 7, Z2
The AVR output is connected to the Exciter Field via terminals marked X – XX or F1 – F2. So
should the Exciter field be megger tested it would be these two leads that must be
disconnected.
A remote ‘Volts–Trim’ adjustment ‘Pot’ is connected across Terminals 1 – 2, if this option is not
being used, then 1 must be linked to 2. A Quadrature Droop CT secondary winding will be
connected to terminals S1 – S2. Terminals A1 – A2 may be connected to a separate excitation
adjustment module, most commonly this second module is a PFC3 [Power Factor Controller,
used when operating in parallel with a Utility / Grid]. If a PFC3 module is being used then this
will be connected to the alternator stator and must be disconnected prior to the megger test.
It may be that Generating Set instrumentation and monitoring equipment is connected directly
to the alternator’s stator output terminals. Care should be taken to check for such connected
equipment and ensure this is disconnected before conducting a megger test. It may be possible
to simply remove fuses to isolate some of these items.
Expected IR Values
The alternator’s Insulation Resistance, along with many other critical factors, will have been
measured during the CGT Factory processes of manufacture, assembly and test. The
alternator will have been transported with a packaging appropriate for the journey to the
Generating Set assembler’s works, where it is expected it will be stored within a weather
protective building.
However, absolute assurance that the alternator will arrive at the Generating Set builder’s
production line with IR values still at the factory test levels of above 100 Mega-Ohms cannot
be guaranteed. The alternator should arrive at the Generating Set assemblers works in a clean
If the unused / new alternators IR values fall below 10 Mega-Ohms, then a drying out procedure
should be implemented by one of the processes outlined in the next sections before being
despatched to the Customers site and some investigation undertaken into the storage
conditions the alternator had been subjected to.
It must be noted that an observed change can be detected with the level of the measured IR
value between the cold and the hot winding assembly condition. It is quite normal to observe
that the ‘hot’ IR value may well be much lower than the unnecessarily, excessively high Ohmic
value measured on the ‘cold’ wound assembly.
When a value of several Mega-Ohms are recorded for the ‘hot’ condition of the winding
assembly, which is several times higher than the minimum value of 1 Mega-Ohm stipulated by
some Marine Classifying Societies, this indicates that the dielectric performance of the
insulation system is still in excellent ‘health’.
This cold-to-hot change of the IR value is associated with the impregnation resin’s internal
molecular stress associated with the wound assembly’s thermal expansion, and the need for
the impregnation seal not to be broken. The chemist played off one characteristic against
another in order to maintain an excellent hot bond strength to assist mechanical rigidity of the
wound assembly in order to combat imposed vibration and forces from external sources such
as; the prime mover, the characteristics of the connected electrical load, and the challenges of
the local operating environment. The chemist has achieved all his objectives, including
ensuring more than adequate dielectric strength is still being provided.
When the alternator is in service, various mechanisms will contribute to factors that will affect
the IR value and site measurements of just a few Mega-Ohms become more typical. Major
factors that affect and reduce the IR value start with the winding outhang. Surface moisture,
often present in conjunction with surface contamination, are both the result of prevailing site
conditions. Either will seriously reduce measured IR values and if the root cause that allows
these contaminants to be present is not addressed, then the expected mean time before failure
(MTBF) will be considerably reduced. The CGT manual - Installation, Service & Maintenance
Manual (Owner’s Manual) - issued with every alternator, in the Service and Maintenance
Section, offers guidance about measuring the IR value and expected typical values.
Ways to Increase IR
Any temporarily reduced IR values can be restored to expected values by adhering to the
following explained drying out procedures. Then a further operation on load will complete a
thorough heating of the winding assembly and will completely dry-off the IR reducing moisture.
Consider a known good condition alternator that has not been run for some time, but during
which it has been standing in conditions of high Relative Humidity (RH). It is possible that
If the above unexcited ‘blow-drying’ does not raise the IR value sufficiently then consider
separately exciting the alternator with its output terminals connected by appropriately rated
short circuit links. By separately-exciting at a level such that some 80% of rated current flows
through the short circuit link the alternator stator winding will start to heat up. This will drive off
the embedded moisture, during a run period dependent upon the size of the alternator but
consider:
If the IR value remains below 1 Mega-Ohm, even after the above drying methods have been
properly conducted, then a Polarization Index test (PI) should be carried out (see next section).
This will offer some guidance about the characteristics and so reason, for the ‘leakage current’
that is causing the low IR.
If surface contamination is the cause, therefore promoting surface tracking, then the alternator
must be removed from the engine and stripped down to enable the stator winding to be
thoroughly, but carefully, pressure washed, then baked in an oven to completely dry out.
If now the value of IR is acceptable, then the stator should be treated with an over-coating
resin that is compatible with the original factory impregnation, before the generator is
reassembled onto the engine.
This test involves using a Megger type instrument that can be set to apply the test voltage to
the winding twice in a period of ten minutes. Usually, a motorised Megger is used. The value
of the windings IR is recorded after 1 minute of the motorised Megger test and then again after
10 minutes. The two values are compared, the 10 minute reading is divided by the 1 minute
reading and hopefully, the result is in the order of 3.
A very basic explanation of a PI test is that it enables an observation of the separation of the
three basic elements that sum together to give the total leakage current identified by a normal
quick Megger test and the correspondingly indicated IR value.
If the molecular structure of the materials in an insulation system are subjected to an electrical
voltage achieved by energisation (using a Megger) of the electrical windings on either side of
the electrical insulation barrier (insulation system of aramid materials and impregnation resins,
etc.), then the insulation material’s molecules will gradually all align with their +ve end facing
in one direction and so their -ve end facing towards the other side of the barrier. When this
situation is complete, Polarisation has been achieved.
Making these molecules align is the result of a capacitance effect across the insulation material
‘barrier’ region. This capacitance ‘charging’ current will cause a Megger instrument to detect
current flow and so indicate a correspondingly low(ish) IR reading.
When a PI test is conducted, the one minute reading is after the Capacitive Current (item 1
above) has occurred, but item 2 and 3 are still 'leaking' and so the IR value is (relatively) low.
When the 10 minute reading is finally taken the Polarisation Current (item 2) is complete and
so now only leakage current (item 3) is present.
The two measured IR values offer guidance about the overall condition of the winding
assembly’s insulation system. If both readings are low, but the ten minute reading is only a
little higher than the one minute reading and a visual inspection confirms a very dirty, surface
contaminated winding, then this assembly could benefit from a thorough washing and drying
process.
Normally a PI test would be conducted on a winding assembly that has been in service, which
visually appears to be clean and dry. But whilst in service, inevitably, the insulation materials
will have been subjected to working dielectric stress, some coil flexing in the out-hang resulting
in minor relative movement between adjacent conductors and the presence of some surface
contamination on out-hangs and through the stator bore.
A PI test on a healthy winding assembly will result in the ratio of the 10 minute reading to the
one minute reading in the order of 3 (or greater) : 1. But an experienced tester will take great
interest in the actual IR values of both readings.
If a PI test is conducted on a new, perfectly clean and virtually unstressed stator winding
assembly, then the one minute reading is usually of such a high IR value, that when the ten
minute reading is taken there is little increase in this IR value over the one minute reading.
Now, when the two readings are ratio’ed, the PI value is often only around 2:1.
Obviously, this does not mean that the new stator assembly is unserviceable, simply that the
PI test has not been a justifiable test to establish the condition of a new stator assembly. Under
such conditions the IR values provide proof of the real condition of the stator assembly.
Basically, keep the wound component clean and dry and good values of IR are assured. But
achieving the ‘clean and dry’ condition has to be engineered into the original Generating Set
design and then maintained by an effective On Site ‘Care regime’.
Considerations at the Engineering stage must include a way to combat moisture and
contaminants carried airborne with the cooling air especially when the Generating Set is
operating in Rain, Fog, or Sea Mist, etc.
Then also carefully consider the microclimate conditions that occur inside a typical Generating
Set canopy, especially just after the hot unit is taken out of surface and is stopped. Within the
canopy the cooling down of the Generating Set promotes high humidity levels, which result in
surfaces become wet, and this includes the alternator windings. It is necessary to find a way
to combat this and basically, it requires engineered natural convection of airflow through the
canopy, to counteract the dew point situation and so keep the levels of Relative Humidity [RH]
low.
Generating Set Canopy Space Heater. Of a power rating capable of maintaining low
Relative Humidity values within the Generating Set enclosure at the site environmental
conditions.
The reference document for the BIL situation is the ANSI / IEEE Std 141 Red Book that is titled
“IEEE recommended practice for Electrical Power Distribution for Industrial Plants”. In
summary the BIL situation sets out to establish the withstand characteristics of the insulation
system when this system is subjected to a temporary excess-voltage stress over and beyond
the normal operating voltage.
The idea of the BIL test is that the insulation should be capable of withstanding a short, sharp
impulse voltage, and sets out to subject the insulation system to a steep fronted, hi-voltage
stress of very short duration. However as the IEEE book points out; “Rotating machines have
Typically, the applied BIL voltage should have a peak [crest] value that is 25% higher than the
standard ‘flash test’ voltage –but remember that the ‘flash test’ voltage is always identified as
V.rms. This BIL [impulse or surge] test voltage is applied from zero to peak in 1.2micro seconds
and then allowed to decay over 50micro seconds.
Whilst Cummins Generator Technologies do not subject their alternators to an actual BIL test,
they do subject them to a ‘Flash Test’, which has characteristics of both a standard ‘flash test’
and satisfies the intent of the BIL test. The applicable standard for rotating elect-machines BS
4999, has guidance about ‘flash testing’ that allows a higher than the standard 2.2kV voltage
to be applied for a shorter period than the standard 1 minute.
Voltage Withstand, High-Pot, Flash Test are all terms applied to the test that involves
subjecting a winding to a test, which applies a mains frequency (ac) high voltage to the winding
groups in turn phase to phase and phase to earth. This test is maintained for a period of
typically one minute, during which no breakdown of the EIS must be detected. This is a test
which should only be undertaken after due technical consideration following other less stressful
test procedures.
Note……. It should be accepted that to undertake a Hi-pot test on an alternator that is in service
is likely to take several hours ( 4<6 ) typically.
This time is based on the restricted access and time taken to gain access to the ancillary
electrical equipment, which must be very carefully disconnected before the Hi-pot test can
begin. The hi-pot test must be undertaken in a restricted access area and prevailing local
legislative health and safety requirements regarding electrical testing must be applied.
One the Hi-pot test is complete the alternator ancillary circuits must be very carefully
reconnected, and the initial run-up of the Generating Set should be undertaken with care and
attention that all is functioning correctly.
The applied test voltage is of a magnitude which is typically three to four times the winding
assembly’s normal operating level and so will severely test the integrity of the winding’s EIS.
If this test is not carefully conducted with due consideration of the damage which can result
from recklessly subjecting electrical systems to inappropriate test voltage levels then EIS
breakdown and associate tracking paths will occur and so permanent damage will be present
within that assembly which will render life expectancy very short.
During the manufacturing process of the alternator, each wound assembly is subjected to a
Hi-Pot test as part of the final inspection function of the QA process.
IEC 60034-1 then advises that following a successful test at the manufacturing QA test point,
the wound assembly should NEVER again be subjected to a Hi-Pot test at that level.
Application Guidance Notes are for information purposes only. Cummins Generator Technologies reserves the right to change the contents of
Application Guidance Notes without notice and shall not be held responsible for any subsequent claims in relation to the content.