TEST DESCRIPTION
It is an off-line, DC test that obtains
information concerning overall status of the insulation. It consists first in
the measure and comparison of the capacitance of the insulation submitted at
low AC and DC voltage, and second in the analysis of the current through the
insulation during two charge and discharge cycles performed at two different
test voltages. In this case, the behavior should be linear when the machine
status is good. If this is not the case, we have an indication of degradation.
EDA test is a non destructive DC voltage Test. The test consists in the
following sequence of data acquiring: •
Environmental Data acquisition:
Ambient temperature and moisture is obtained.
This will be part of the report. If the machine was switched off long enough,
this temperature could be used to compensate the parameters read. Moisture is
used to have a useful indication of ambient conditions when the test was
performed, both for air cooled machines and to know if measurement
instrumentation performance is exceeded.
• Capacitances measurement:
The
system performs insulation capacitance measurement at low voltage and at two
frequencies: DC and 1KHz AC. •
First
Test Voltage:
System
switches DC test voltage to the insulation while 30 minutes. Currents are
plotted and recorded meanwhile to determine several parameters. At the end of
the 30 minutes, the system performs a 2 minutes discharge period.
Second Test Voltage:
After
complete discharge of the insulation from 1st test voltage, a second test
voltage with a value above the 1st is applied to the machine. The procedure is
the same. This step is made to check for possible problems in insulation at a
high voltage and second to check about linearity of insulation with voltage.
Diagnosis:
The
diagnosis management works in two steps: First and as reference of initial
values, an electrical engineer performs and EDA test during machine
commissioning or after any refurbishment, stating so that the reference values
are those of health insulation. After that and during routine maintenance, once
finished the test, the inspector can immediately check the numeric values of
all the parameters and their comparison with previous test done and the initial
reference values and have so that a punctual view of the status of the machine
insulation.
Trending:
Predictive
maintenance techniques need of trending data that can help us decide about last
parameter’s evolution and indicate the future tendency. With this data,
inspectors can decide about future maintenance actions on the machine…
2.- TEST PARAMETERS AND INTERPRETATION
EDA
system offers a group of parameters that when analyzed together give a very
good approach of insulation status. Now we indicate each parameter, what does
the parameter indicate and how it changes because of different problems with a
simple explanation. The main parameters we analyze are: • Capacitance Ratio
(CR). This is the ratio between DC and AC (1KHz) capacitances. This ratio
allows detect contamination and moisture problems.
Contamination:
Contamination
extends all over the windings in the end-winding area. This contamination
extends the plates of the equivalent geometric capacitor formed between the
phases and the magnetic core. This extension is measured in DC, but in AC part
of the capacitance is neglected due to the effect of the inductance of the
coils, so the bigger the difference, or the greater CR, the greater the
contamination degree.
Moisture:
When
moisture is in the surface of the insulation, the effect is similar to
contamination. If moisture is inside the insulation, then it is the dielectric
constant of the insulation that changes from the value in dry insulation
increasing capacitances.
Aging:
The
insulation may degrade simply by the pass of time or by another accelerated
process usually due to thermal aging. In both situations, components inside the
insulation volatilize and the space is filled by air or another gases. In this
condition, the effective distance between layers of the capacitor decrease, and
also the effective dielectric constant of the equivalent capacitor decrease.
Both situations lead to a low but noticeable annual decrease of the machine
capacitances that can be easily detected with data trending. •
Absorption
Index (AI).
It is a ratio between the currents after 30
and 60 seconds subtracting the leakage current while the 30 minutes charge
cycle. It indicates if a problem is internal (low values) or external (high
values) to the dielectric. It is based in changes of the ground wall insulation
homogeneity due to the presence of contaminants coming from the outside:
moisture, oil or others ingresses flowing into the dielectric.
This
behavior can be appreciated again from figure 7 because: Cg gets charged before
the 30 seconds and in 30 minutes leakage current is most contributed from Ri,
so Ri influence is neglected. What remains?, only the effect of the equivalents
Ra/Ca and Cwx., If the plot of the current gives a defined corner slope it gives
high AI values indicating external contamination, if the slope is slow it means
that the insulation has problems in the Ra/Ca component indicating through low
AI values internal problems. •
Re absorption Current (RC).
After
30 minutes, the insulation is considered fully charged. Then a short-circuit is
made and flows the Re-turned absorption current that was acquired while the
absorption period but this time without the influence of the other currents.
The value is taken after one minute of the discharge cycle. If there is a
bigger heterogeneity inside the dielectric (water, vapor, oil, dust…), it has
more areas to store energy as polarized dipoles, leading to bigger
repolarization currents. From previous tests, if RC goes up, it is possibly due
to the degradation of the insulation (delimitation, increase of voids, etc).
It’s possible to detect uncured resin if we compare this parameter with test
results from similar machines. •
Time Constant (TC).
It is
the product of an Insulation Resistance (IR) value and the AC Capacitance. New
machines must show high initial values because of the high IR expected initial
values. In a normal life and degradation, both IR and CAC decrease, so
throughout the life of the machine TC descends slightly. Therefore, if the value
falls faster than expected, it indicates a faster aging, and so an abnormal
degradation. It confirms contamination or delamination problems, depending of
the case. It must be temperature normalized. •
Leakage
Current Ratio (LCR).
It’s the ratio between the leakage current
(current after the 30 minutes of charge process) measured at the first voltage
and at the second voltage. The result is made comparable through the voltage
ratio between the two processes. The ideal value is 1,0: This value would mean
that the insulation behavior is linear with voltage. If the value is far from
the ideal value, it should confirm a problem detected with the other
parameters. •
Insulation
Resistance (IR).
Every machine should be above a minimum value,
depending on the machine voltage. The decrement of this value indicates
insulation contamination or degradation. After external cleaning it should
increase noticeably. As can be seen in figure 7, Ri is made up from surface
resistance influenced by contamination and internal resistance that includes
internal insulation contamination and possible aging. It requires temperature
correction. •
Polarization
Index (PI).
It
indicates together with IR different problems like contamination and moisture
absorption. If PI tendency is going down from previous tests, the machine has
one of the described problems. Otherwise, if PI increases in excess, it can
indicate brittle insulation in some older types of insulation
3.- SECOND INSTANCE STATOR INSULATION TEST
TECHNIQUES
When
EDA test gives evidences of one or more insulation problem, it’s the time to
confirm this issue with 2nd instance techniques, before taking the decision of
opening the machine. These techniques usually need more expensive or complex
equipment and needs better trained people not just for the testing but for the
interpretation of the results as well, so they are only performed when the
first instance tests recommends a deeper study. These techniques are also non
destructive tests and could be:
•
Internal Visual Inspection.
The
scope of this inspection will depend on the accessibility to the winding of the
machine, performed by expert personnel. This will be the most valuable tool to
detect the effects of the different degradation mechanisms, especially if those
happen in the end-winding area. •
Power
factor or Dissipation factor (tan δ) and tip-up.
Tan
delta gives a global indication of losses in the dielectric and can give a clue
of the thermal aging status of the machine. Also, this test could give
information about end-winding area with UST Test Techniques. •
Capacitance and capacitance tip-up.
High voltage capacitance deduced from tan δ
test also gives clear indication of changes in the insulation system. It
monitors moisture and other abnormal conditions. The tip-up indicates how
changes the equivalent capacitance with voltage, this says us how changes the
amount of PD (Partial Discharges) voids activated by voltage increase. It also
monitors moisture ingress and delamination. •
Partial
discharge analysis.
Monitors degradation of the insulation system.
It can be on or off-line. It can also give an indication of the possible
location of the discharges inside the dielectric (next to the coil, in the
ground wall or in the interface to the core). •
TVA probe.
TVA
Probe testing (Tennessee Valley Authority Probe) allows physical location of
the partial discharge activity, mainly if it is a slot discharge problem. •
Surge Test.
It is intended to test turn-to-turn insulation
in form-wound multi turn coils. •
Hi-pot.
It is
a destructive test that should be performed with extreme caution. It must be
evaluated the convenience of the performance of this test.
4.- TEST DESCRIPTION
Here
we present a study of some real cases to show how the value and the trends of
these parameters are reliable and help to detect evolving problems in several
situations. The cases are taken from Iberdrola testing files and describe
problem detection through different EDA test results and how the described 2nd
instance techniques confirm the problems.
All
test performed were validated before being processed. Test system (EDAIII) is
intended to eliminate most of human error, but apart from this, many test were
invalidated because of very high moisture in ambient or machine. This could
cause moisture deposit in machine’s surface, most of failed tests were repeated
after the machine was dried and results changed to normal values. After
validation, and although more complex intermediate results could be defined,
for this study, tests diagnosis are only classified in four categories, the
most usual in the test environment and corresponding to the records on machine
maintenance history:
[1]
Healthy machines, suitable for operation.
The
insulation is in good condition, there is no serious contamination level. If
the unit is old (>30years), the degradation degree is according to normal
aging and doesn’t show symptoms of accelerated or abnormal thermal aging.
[2]
Surface contamination problems.
Due
to dust, insects, oil or simply moisture or also to a combination of several of
these factors. If nothing was done, it could lead to electrical tracking and
contamination ingress in the insulation decreasing its useful life. It is most
of the times easily reversible with cleaning & drying works. 10
[3] Internal degradation problems.
Usually
due to thermal aging. It is irreversible and makes necessary a more intensive
trending up to the rewinding of the machine be needed. Insulation exceeds usual
aging values; it is losing its properties and can support less stress. The
machine is subject to a greater probability of failure in service.
[4] Internal contamination problems.
This situation is not very usual in epoxy
insulation, but sometimes and due to cracks or fissures in insulation,
delamination and ingress of moisture, dirty or oil can also flow inside the
coil insulation. It is most of the times hardly to reverse.
It is obvious that some machines can have
(normally) more than one problem at a time. For this study, only one problem is
allowed in each diagnose, the usual preference is: First Health machines,
Contaminated & Degraded as the last. Take note that both contaminations can
change Degraded values and change Degradation diagnosis, so the user should
first try to solve this easy to reverse problems in Contaminated machines before
evaluating other hardly to solve problems related to Degradation. It is also
logic that most of the times that Internal Contamination happens; it will also
show External Contamination traces.
The
last step of the investigation was to execute histograms for the different test
parameters under study: For each parameter, the first histogram presents data
distribution from low voltage / high voltage test. The second histogram
presents data classification for this four diagnosis categories.
PROCEDURE:
THE
STEP VOLTAGE TEST CONSISTS IN THE INSULATION OF HIGH DC VOLTAGE ACROSS THE
INSULATION (ALWAYS ALLOWS MACHINE’S OPERATION VOLTAGE) LASTING 30 MINUTE
FOLLOWED 30 MINUTE TIME.
PROCESS
FOLLOW TO EACH BUSBAR SAME AS BELOW.
1ST
GENERATOR
R-PHASE àGENERATOR
EARTH
GENERATOR
Y&B-PHASE SHORT à
EARTH
2ND
GENERATOR
Y-PHASE àGENERATOR
EARTH
GENERATOR
R&B-PHASE SHORT à
EARTH
3RD
GENERATOR
B-PHASE àGENERATOR
EARTH
GENERATOR
R&Y-PHASE SHORT à
EARTH
EQUIPMENT:
MEGGER
MIT S20/2 SKV INSULATION TESTER
RESULT:
Temp: 38°C
DC Voltage in kV |
Leakage
Current in µA |
||
R-Phase |
Y-Phase |
B-Phase |
|
1.6 |
1.30 |
1.75 |
3.74 |
3.2 |
2.87 |
3.18 |
7.44 |
4.8 |
4.55 |
4.66 |
11.2 |
6.5 |
6.58 |
6.42 |
15.5 |
8.1 |
8.85 |
8.29 |
20.0 |
0 Comments