INDUSTRY TOPICS
TESTING ROTATING MACHINERY
POLARIZATION/
DEPOLARIZATION
CURRENT (PDC)
B Y V I CK I WA R R E N ,
Iris Powe r LP
The insulation resistance (IR) and polarization index (PI) tests [IEEE
Std. 43-2000, ANSI/NETA ATS/MTS], should be done prior to
application of any high voltage tests or return to service to assure
that the winding is not wet or dirty enough to pose a risk of failure
that might be averted by a cleaning and drying-out procedure.
IR/PI is a useful indicator of contamination
and moisture on the exposed insulation
surfaces of a winding, especially when there
are cracks or fissures in the insulation. These
tests are easily done (see NETA World Winter
2011 and Spring 2012 issues). Since squirrel
cage induction motor rotor windings are not
insulated, these tests are not appropriate.
Resistance testing is principally a pass/fail
criterion and cannot be relied upon to predict
the condition of the main insulation except
when the insulation has already faulted.
Experience has shown that IR/PI is a useful
indicator of contamination and moisture on
the exposed insulation surfaces of a winding,
especially when there are cracks or fissures in
the insulation. However, as discussed in IEEE
43, the IR/PI test does not seem to be sensitive
to many other stator winding insulation
problems such as:
•Loose coils in the slot that lead to
insulation abrasion
•Delamination of the insulation due
to operation at high temperature
94
•
SPRING 2014
•Separation of the copper from the
groundwall insulation due to load cycling
• Deterioration of the stress relief coatings
•Partial discharge (PD) between coils in
different phases due to insufficient
spacing in the endwindings
AC tests, such as partial discharge (described in
NETA World Winter/Spring 2013) are effective in
finding these issues, but are cumbersome to do in
an offline configuration due to the need of a large
ac power supply; therefore, more sophisticated
dc tests that some have proposed may detect
more kinds of problems than the simple IR/PI
test. Included in the NETA World Spring 2012
issue was a section about the dielectric response
analysis (DRA) or polarization/depolarization
current-measurement (PDC). This measures
the charging and discharging currents of the
winding insulation of stator or rotor winding.
Reportedly, the results of the measurement
provide information about the condition of
machine insulation (cleanliness, humidity,
ageing, corrosion, resin decomposition, and
similar characteristics).
TESTING ROTATING MACHINERY POLARIZATION/DEPOLARIZATION CURRENT (PDC)
INDUSTRY TOPICS
THEORY
The new draft standard (IEEE Std. 43 draft
revision 2012) has an annex that deals with
additional information that can be obtained
by applying a stable dc voltage to a complete
stator winding or individual phases for
1000 to 2000 seconds and recording the
polarizing current IP versus time. The
voltage is then removed and the discharge
current ID is monitored as a function of
time using a suitable discharge circuit.
When the voltage is removed, reverse current
flows and the molecules in the insulation
become disorientated and the space charge
dissipates. This discharge current ID
has two main components: a capacitive
discharge current component, which decays
nearly instantaneously, depending upon the
discharge resistance; and the absorption
discharge current, which will decay from
a high initial value to nearly zero with the
same characteristics as the initial charging
current but with the opposite polarity.
Normally, neither the surface leakage IL nor
the conduction current IG affects the discharge
current. Differences in the IP and ID [Figure 1]
may indicate winding lack of curing, moisture
absorption, surface contamination, damage to
the voltage stress coatings, or severe thermal
deterioration of the bulk of the insulation.
Figure 1: Charging and Discharging Curves
EXAMINATION OF THE PDC TEST
Since the IEEE 43 standard suggests that the
PDC test is valid for testing windings for
thermal deterioration, tests were done on a
motor stator rated 13.2 kV, 6000 HP known to
be thermally aged, but clean and dry. Both the
PDC test and the ac offline partial discharge
test were done, as the latter has proven to be
effective for evaluating insulation delamination.
The stator had an asphaltic mica insulation
system and was several decades old. The three
Figure 2: PDC plot for A-phase with a 10 kV charge cycle (other phases were identical). The charge
current is the upper line and the discharge current is the bottom line. The logarithmic vertical scale goes
from 0.1 μA to 1000 μA and the horizontal axis ranges from 1 to 1000 seconds.
TESTING ROTATING MACHINERY POLARIZATION/DEPOLARIZATION CURRENT (PDC)
NETAWORLD
•
95
INDUSTRY TOPICS
phases in the stator could be isolated from each
other to facilitate testing of each phase.
Figure 3: LF PD plot for phase A, which had the highest PD at 8 kV.
Note that the polarity of the PD plotted between 0 and 180 degrees has
been inverted. The linear scale ranges from 0 to 11 nC.
Figure 4: PD response from stator phase A, which had the highest
PD at 8 kV.
96
•
SPRING 2014
Polarization/Depolarization Test
For these tests, a PDTech DRA 3 was used
to record the polarizing and depolarizing dc
currents. It applies a positive dc voltage to
the test object at a selected voltage (usually
5 or 10 kV for these experiments) and for a
selected time (usually 1000 s – about 16.7
minutes), while measuring the charging
current. The dc supply is then removed
from the test object and the test object is
grounded. The discharge current-to-ground
is then measured for the same amount of
time. Software records these currents, inverts
the discharge current, and displays both the
charge and discharge currents
in the same plot against time, with the plot
time origin starting from either the start of
the charge cycle, or the start of the discharge
cycle (Fig. 2). The difference in the charge
and absolute value of the discharge current
can also be displayed. The instrument also
calculates the IR and PI. All tests were done
at 20°C.
Partial Discharge Test
Partial discharge tests were conducted using
a conventional IEC 60270 PD detector, a
PDTech DeltaMaxx, working in the low
frequency (<3 MHz), broadband range.
This instrument automatically converts
the measured pulse magnitude in mV into
pC, per the procedure in IEC 60270. The
PD was also measured in the VHF range
(30-300 MHz) range with an Iris Power
TGA-B, more typically used for on-line PD
monitoring. This device does not perform an
automatic normalization, instead it reports
the PD magnitudes in mV. The PD tests were
performed at rated line-to-ground voltage,
and the data was recorded after stabilization
at the test voltage for 10 minutes. The low
frequency (LF) and VHF phase-resolved PD
plots for the worst phase (A-phase) are shown
in Figures 3 and 4, respectively.
TESTING ROTATING MACHINERY POLARIZATION/DEPOLARIZATION CURRENT (PDC)
INDUSTRY TOPICS
Table 1: Summary of Diagnostic Results for the Motor Stator Winding
Test Configuration
IR
PI
(GΩ)
VHF PD (mV)
LF PD (nC)
DF (@2 kV)
Tip-Up
Qm+
Qm-
Qm
(%)
(%)
A
5.1
3.4
1066
888
9.7
4.7
0.79
B
5.2
3.4
275
330
3.7
4.6
0.73
C
5.0
3.4
433
440
2.9
4.8
0.80
Conclusion
Table I shows a summary of the test results
for each phase of the winding. PDC plots
for each phase are shown in Fig. 2. When all
three phases are super-imposed, all the charge
currents and all the discharge currents overlap
completely.
It is clear from Table I that A phase has the
highest PD in both frequency ranges. Since
the positive and negative PD is about the
same (within +/- 25%), then based on normal
PD interpretation rules (see IEEE 1434
or IEC 60034-27), one suspects the PD is
due to groundwall delamination, and when
compared to the instrument manufacturer’s
database would be ranked as Very High.
PDC has been proposed as an off-line tool
that can detect issues besides contamination
and moisture absorption, such as something
that can be used to confirm the diagnosis of
insulation condition obtained from on-line
PD testing, without having to use a large ac
transformer to energize the winding for an
off-line PD or tip-up test. The comparison
tests described here did not produce consistent
results. Clearly more tests are required to
determine its efficacy.
REF EREN C ES
1. IEEE, Recommended Practice for Testing Insulation Resistance
of Rotating Machinery, IEEE std. 43-2000
2. ANSI/NETA, Standard for Acceptance Testing Specifications
for Electrical Power Distribution Equipment and Systems, 2009
edition
3. ANSI/NETA, Standard for Maintenance Testing Specifications
for Electrical Power Distribution Equipment and Systems, 2007
edition
4. Stone, G., and Sasic, M., Experience with DC PolarizationDepolarization Measurements on Stator Winding Insulation,
Electrical Insulation Conference (EIC), Ottawa, Ontario,
March 2014.
5. IEC 60034-27, “Off-line partial discharge measurements
on the stator winding insulation of rotating electrical
machines”
Vicki Warren, Senior Product Engineer,
Iris Power LP. Vicki is an electrical
engineer with extensive experience in
testing and maintenance of motor and
generator windings. Prior to joining Iris in
1996, she worked for the U.S. Army Corps
of Engineers for 13 years. While with the
Corps, she was responsible for the testing
and maintenance of hydrogenerator windings, switchgear,
transformers, protection and control devices, development of
SCADA software, and the installation of local area networks.
At Iris, Vicki has been involved in using partial discharge
testing to evaluate the condition of insulation systems used in
medium- to high-voltage rotating machines, switchgear and
transformers. Additionally, she has worked extensively in the
development and design of new products used for condition
monitoring of insulation systems, both periodical and continual.
Vicki also actively participated in the development of multiple
IEEE standards and guides and was Chair of the IEEE 432000 Working Group.
TESTING ROTATING MACHINERY POLARIZATION/DEPOLARIZATION CURRENT (PDC)
NETAWORLD
•
97