Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
POWER2008-60167
James E. Timperley P. E.
IEEE Fellow
Doble Engineering
Columbus, Ohio, USA
ABSTRACT
Photo 2
Connection to Collect Generator EMI Data
INTRODUCTION
Microvolts (Quasi-Peak)
Corona
100
12-18-00
stages.
10
10-05-05
1
Pd's &
Corona
0.1
0.01 0.1 1 10
Frequency (MHz)
Curve 2
Motor C after Repairs
Frequency (MHz)
10000 4000 volt 4000 hp
Circ. Water Pump Motors
Curve 3
1000
Microvolts (Quasi-Peak)
SCR noise
Motor “B” Has Broken Rotor Bars
Slip noise &
process noise Partial discharges
100 in motor lead box
Sleeve bearing problems and broken rotor bars can be
10
expected during the service life of large induction motors.
CWP Motor B Motor “B” in Curve 3 has developed both problems. A hand
1 CWP Motor A held EMI detector is used to isolate the location of each defect.
SCR noise
Motor “A” has no requirement for maintenance.
0.1
0.01 0.1 1 10
Frequency (MHz)
Curve 1
EMI Signatures of Two CWP Motors
1000
100 AM
Microvolts (Quasi-Peak)
Exciter SCR stations
noise
* Random
10
Corona 100 * noise
Resonance
Gap
Noise
1 2 Middle Condensate Booster 10 11-05-01
2 North Condensate Booster
12-03-03
0.1
0.01 0.1 1 10 1 * Developing wet stator bar
Frequency (MHz)
0.1
Curve 4 0.01 0.1 1 10 100
Comparison of Clean and Dirty Motors Frequency (MHz)
Curve 5 10
This Machine Has Loose Stator Bars
1 Exciter
noise
Tightening a generator stator winding can be expected
during the normal service life. Curve 5 is an example of a 0.1
generator power down test. When a stator winding is tight 0.01 0.1 1 10 100
changes in the load will not change the EMI signature. If a
stator winding is loose then EMI amplitude follows the load. Frequency (MHz)
The generator above has a reduction of EMI mean levels as the
load is reduced. This indicates the stator winding is loose and Curve 7
should be tightened. However the question was could Minor Shaft Ground Sparking
rewedging be postponed for two cycles while major turbine
work was performed? The answer was yes, rewedging could be The generator in Curve 7 has minor sparking at the shaft
postponed with minimum risk. The activity at full load did not grounding brushes. Carbon brushes tend to glaze over and are a
indicate severe looseness was present. Rewedging was high maintenance option for shaft grounding. Cleaning the
postponed to a later outage and has recently been completed.
Microvolts (Quasi-Peak)
100
11-10-04 NCC Motor B
Slip noise Moderate corona
1,200 MW Generator with severe arcing Test cables
10000 of shaft grounding braids 10 "B" Motor
Microvolts (Quasi-Peak)
1000 Power
Line
With Arcing 1
Carriers
Shaft Grounding Braid
100
11-10-04 NCC Motor C
10 0.1
0.01 0.1 1 10
W/O Arcing
1 Frequency (MHz)
0.1
0.01 0.1 1 10 100 Curve 10
Frequency (MHz) Defective 4 kV Power Cable
Curve 8 There have been a few cases where power cable problems
Severe Shaft Grounding Problems have been detected while testing a motor. In Curve 10 the “B”
motor has high cable corona levels. There should be no
The generator in Curve 8 had a new shaft grounding measurable partial discharge activity on shielded 4 kV cables.
system installed. The arcing was severe after a few weeks Additional cable testing was recommended.
operation. Equipment misoperation resulted from the extremely
high levels of EMI that were generated. A new shaft grounding
assembly was recommended.
1000 Noise
spike
100
Exciter noise Random
noise
10 Div 2 Diesel Gen.
11-10-04
Div 1 Diesel Gen.
1 09-08-04
0.1
0.01 0.1 1 10
Frequency (MHz)
Curve 9 Curve 11
Signatures of Emergency Generators Wet 4 kV Power Cable
Curve 9 shows the comparison of two emergency diesel In Curve 11 motor “B” has no indications of any
generators. Data was collected during the routine load test. problems, the local AM radio stations have higher levels than
Data was collected from a generator frame ground since the system related PD activity. However the cable tunnel for the
winding neutral was not accessible. No problems were “A” motor is flooded and there are high cable corona levels
expected and none were detected. Most of the spectrum is across the spectrum.
dominated by the exciter power electronics. If winding
problems were present they would be seen at frequencies above
1 MHz.
Hydrogen
Quasi-Peak (microvolts)
Isolated phase bus Seal Rub
discharges
Stator related activity
100 Exciter
October, 1996 Diode
10 Transients
10
June, 2001
1
August, 2002
1
June, 1996 Microsparking
in Isolated Phase Bus
0.1 0.1
0.01 0.1 1 10 100 0.01 0.1 1 10 100
Frequency (MHz) Frequency (MHz)
Curve 12 Curve 14
Developing Isolated Phase Problems An Exciter End Seal Rub
In generator in Curve 12 developed severe isolated phase During an annual EMI Diagnostic the generator in Curve 14
bus discharges after bus maintenance was performed in 1996. had an increase in activity around 8 MHz. The hand held EMI
EMI levels at frequencies above 10 MHz had increased. A bus detector located the source as the generator exciter end
inspection and review of past activities was recommended. hydrogen seal. The seal was replaced during the next outage.
There was severe pitting on the brass seal from the shaft
currents that had caused the high EMI levels.
1400 MVA 24kV 1800r/min
Generator
1000
345 / 13 kV
Quasi-Peak (microvolts)
Start-up Transformers
Isolated Phase
100 Gap Discharges
Comparison
October 9, 1996
PL AM
Carriers Stations
10 1000
Microvolts (Quasi-Peak)
Curve 13 0.1
Before and After Bus Repairs 0.01 0.1 1 10 100
Frequency (MHz)
The generator shown in Curve 13 is at the same location as
the generator in Curve 12. In 1996 there were only minor bus Curve 15
discharges noted around 30 MHz. These disappeared when the A Defective 345 kV Disconnect Switch
system was retested nine months later. No maintenance was
recommended. Transformer hardware problems can also be detected with
EMI. In Curve 15 the partial discharge source was a loose
connection on a high side disconnect switch of the #2
transformer. This was verified with an ultrasonic scan and
repaired during the next outage.
Arcing
Turbine/Generator Workshop and Vendor Exposition,
1998
1999 August 2003, Nashville, TN.
10
[4] Allen, J. D., and Timperley, J. E., “Scheduling Power
Generator Activity
Plant Maintenance on EMI Diagnostics,” Coal Gen, August
2003, Columbus OH.
1
CONCLUSION
EMI Diagnostics is an expert based technology that can be MOTOR CONDITIONS IDENTIFIED WITH EMI
an important part of a nuclear plant condition based
maintenance program. A wide variety of equipment problems 1. Dirty stator windings
can be detected. Data can be trended to track deterioration rates 2. Loose windings in slots and end-arms
or to measure success of repairs. The power electrical systems 3. Broken rotor bars
are monitored under normal load conditions with no 4. Synchronous motor field ground
interruption to operations. No design changes are needed to 5. Rotor not set on magnet center
collect data. 6. Frame had loose foundation (soft foot)
7. Wiped sleeve bearings
8. Defective outboard bearing insulation (shorted)
ACKNOWLEDGMENTS 9. Bearing oil seal rub,
10. Exciter drive shaft weather seal rub
The author appreciated the help of John Allen, George 11. Coupling misalignment with driven gear box, pump, fan
Karayianopoulos, Stan Carter, and Dave Klinect for their 12. Defective or missing coupling insulation
assistance in the preparation of this paper. 13. Circulating currents in driven pumps
14. Magnetized gear box shafts / gears
REFERENCES 15. Loose neutral connections
16. Loose crimp / bolted line connections
[1] Timperley, “Generator Condition Assessment Through EMI 17. Loose surge / power factor capacitor connections
Diagnostics,” ASMI paper PWR 2008-60166, ASME Power 18. Abrasive erosion of stator windings
2008, July 2008, Orlando, Florida. 19. Defective motor lead insulation
20. Detect wet power cables
[2] Timperley, J. E., Michalec, J. R., Moore W., Moudy, H., 21. Detect 13 kV cable stress cone deterioration
Hutt, J., “Rewind and Core restack of Conesville #5 22. Verify correct maintenance was or was not performed
Generator,” ASMI paper PWR 2004-52145, ASME Power
2004, March 2004, Baltimore, MD.