Download Document 8910228

21, rue d’Artois, F-75008 PARIS
http : //www.cigre.org
A 1-203
CIGRE 2006
EXPERIENCE WITH CONTINUOUS ON-LINE PARTIAL DISCHARGE
MONITORING OF HYDROGENERATOR STATOR WINDING INSULATION
G.C. STONE*, B. LLOYD, M.SUSNIK –
Iris Power Engineering
Canada
R.HUBER
B.C. Hydro
Canada
S. ZELINGHER
New York Power Authority
USA
SUMMARY
On-line partial discharge (PD) testing has become a well-established tool for assessing the
condition of the stator winding electrical insulation. PD tests find stator winding insulation
deterioration problems such as loose coils, contamination and long-term overheating.
Historically, hydrogenerators have been equipped with permanently installed PD sensors, and
readings are taken periodically using a portable test instrument. However many utilities now
see a need to continuously monitor the PD activity. Continuous monitoring, together with a
convenient interface provides many advantages to the user. The data is available on a plant
computer and can be downloaded anywhere. Tests to be done remotely, and better trends of
PD over time under the same generator operating conditions can be produced. In addition,
continuous monitoring enables the maximum possible warning time of stator winding
problems, as well as facilitating over-all generator condition assessment using emerging
expert monitoring systems.
A relatively inexpensive continuous PD monitoring system called HydroTrac has been
developed jointly with the Electric Power Research Institute and the New York Power
Authority and seen operation since 2001. Over 150 hydrogenerators are now equipped for
continuous monitoring using this technology. Since each utility has a preferred way of
passing data between a monitor and the plant computer, several interfaces have been
developed including analog outputs via 4-20 mA links, RS 485, Ethernet, PI and OPC.
Examples of the use of the continuous PD monitoring systems to establish the condition of the
stator winding insulation in two utilities are presented. These examples show that continuous
monitoring is practical, and maintenance plans can be established based on information from
the monitors.
KEYWORDS
Stator winding, insulation, partial discharge, hydrogenerator condition monitoring, automation
*[email protected]
INTRODUCTION
Over the past 15 years on-line partial discharge (PD) monitoring has become the most widely applied
method to determine the condition of the electrical insulation in hydrogenerator stator windings rated 6
kV or more. Partial discharges (sometimes also known as corona) are small electrical sparks that
occur in deteriorated stator winding insulation systems. Over 50 years of experience with on-line PD
testing shows that it will detect most (but not all) of the common manufacturing and operations-related
problems, including:
•
•
•
•
•
•
•
Loose coils in the slot (slot discharge)
Overheating (long-term thermal deterioration)
Winding contamination by moisture, oil, dirt, etc.
Load cycling problems
Poor resin impregnation during manufacture
Inadequate electric stress control coatings in the slot or endwinding
Insufficient spacing between coils in the endwinding area.
Problems such as endwinding vibration or metallic debris in the machine will often not produce PD as
a symptom.
There are many methods available to measure the PD activity in operating hydrogenerators [1-3]. The
electrical techniques rely on monitoring the current or voltage pulse that is created whenever a partial
discharge occurs. Although off-line PD testing is often done as a quality assurance test by
manufacturers and rewinders [4], owners of hydrogenerators prefer an on-line PD test, i.e. one t hat can
be performed during normal operation of the hydrogenerator, since no test outage or large AC supply
is needed. The earliest on-line PD measurement methods detected the PD pulse currents by means of a
high frequency current transformer at the neutral point [2,3], but today most machines around the
world use 80-pF high voltage capacitors on the output bus to measure the PD [1,5].
The majority of on-line PD measurements on hydrogenerator stator windings are done with a portable
test instrument, which is connected to the permanently installed PD sensors for perhaps 30 minutes,
two times per year. However, there are situations where continuous on-line PD measurement may
offer advantages to hydrogenerator owners. This paper outlines these situations, describes the
continuous monitoring systems that have been developed and provides some case studies after a brief
review of PD measurement technology.
CONVENTIONAL ON-LINE PD MEASUREMENT TECHNOLOGY
A particular challenge with PD measurements is encountered when the hydrogenerator is operating
normally. Since the machine is connected to the power system, electrical interference (noise) is often
present. Noise sources include corona from the power system, slip ring or commutator sparking,
sparking from poor electrical connections, and/or power tool operation. This electrical noise obscures
the PD pulses, and may cause the technician to conclude that a stator winding has high levels of PD,
when it is actually the noise. The consequence is that a good winding is incorrectly assessed as being
defective, meaning that a false alarm is given suggesting that the winding is bad, when it is not. Such
false alarms reduce the credibility of on-line PD tests, and even today, many feel that on-line PD
testing is a ‘black art’ best left to specialists.
Twenty-five years ago, the North American utility industry (via the Canadian Electrical Association)
sponsored research to develop an objective on-line PD test for machines that could be performed and
interpreted by plant staff with average training [1,5]. The type of PD test emphasized separating PD
pulses from electrical noise pulses. For hydrogenerators, 80 pF capacitors detect the PD, while
blocking the high AC voltage. In these machines noise separation methods depend on digitally
[email protected]
2
comparing the time of pulse arrival between a pair of 80 pF capacitive couplers on each phase and/or
analyzing the shape of the individual pulses [1,5]. To maximize the signal-to-noise ratio, and thus also
to reduce the risk of false indications, the sensors detect the PD at frequencies of 40 MHz and higher.
These sensors, together with an associated portable instrument called a PDA, have enabled utilities to
assess the winding condition with their own staff. It is estimated that over 50% of all North American
hydrogenerators rated 20 MW or more have been permanently equipped with the required PD sensors.
Globally, well over 6000 machines have the required PD sensors. Interpretation of the PD results is
done by analyzing each test result and trending the stator PD over time. Experience shows if the PD
magnitude doubles, under the same operating conditions in 6 months or so, the rate of insulation
deterioration is increasing significantly [1,4]. More recently, based on a statistical analysis of data
from >60,000 tests, tables of what constitutes a high reading as a function of generator ratings and PD
detection system have been published [1,6].
CONTINUOUS ON-LINE PD MONITORING SYSTEMS
Rationale
Stator winding insulation deterioration is a relatively slow process in which the time between when
significant PD can be detected and when winding failure may occur, is usually two or more years.
Thus, periodic measurements done once every six months or so are often adequate for detecting stator
winding problems with sufficient warning to conveniently implement corrective action. However,
there are several situations where continuous monitoring may be advantageous:
1. Expert systems are being deployed which continuously monitor all the sensors in a generator (for
example temperature, vibration, air gap, current, voltage, etc.) to determine if any problems are
occurring anywhere in the generator [7,8]. This provides plant operators as well as maintenance
personnel with warnings that a problem may be developing, and also gives advice on corrective
action. Since the partial discharge activity is an important source of information with regard to the
stator winding condition, it is desirable that generators equipped with such expert systems have
ready access to PD data.
2. The stator winding PD activity is often affected by hydrogenerator operating conditions such as
winding temperature, load and voltage, as well as environmental humidity [1]. For example, if a
winding is loose (i.e. the stator coils can vibrate in the slot), an increase in generator load will
cause the PD activity to dramatically increase. Although these operating-condition-dependent
results help in interpreting the type of deterioration occurring, they often make it difficult to trend
PD data over time. To obtain trendable data, which is critical for interpretation, the machine must
be tested under the same load, temperature and voltage conditions. For units that are subject to
frequent load cycling, or pumped-storage generat ing units, it is often difficult for the plant
operators to recreate exact operating conditions from test to test. Continuous PD monitoring
solves this problem by continuously measuring the motor/generator operating conditions, and then
storing the PD measurement at which the proper operating conditions occur naturally during
normal operation of the machine.
3. Many hydro generating plants are located in remote areas, making it is expensive to transport test
personnel to the site for doing a conventional PD et st. Furthermore, many hydro generating
stations are remotely operated, that is, there is no permanent on-site staff. A continuous monitor
can allow the PD measurements to be made remotely, without the need for sending personnel to
the plant. This will reduce overall operation and maintenance costs.
4. Sometimes periodic on-line PD monitoring has identified a stator winding that has a high risk of
failure, but operating considerations prevent an immediate repair or rewind. By installing
continuous PD monitoring, plant maintenance and operating personnel can keep a closer watch on
[email protected]
3
the stator winding, possibly extending the winding life with a lower risk of a catastrophic inservice failure.
For the above reasons, first generation continuous PD monitoring systems were developed for
hydrogenerators, and after some teething pains, were successfully deployed [9-12]. However, these
early systems had a relatively high installation cost and did not enable generator operating data to be
extracted the plant computer. Under a joint development project between the USA -based Electric
Power Research Institute and the New York Power Authority, a next -generation continuous monitor
called HydroTrac was developed.
Hardware
The new continuous monitor employed exactly the same noise separation methods that had been used
for many years for periodic monitoring. Also, great care was taken to make sure the same PD
quantities such as Qm (peak PD magnitude) and NQN (essentially the integrated PD activity) were
produced, to facilitate seamless trending with past data obtained from portable instruments. A block
diagram of the hardware is shown in Figure 1.
A key difference between the new system and past continuous monitors is how the generator operating
and environmental data is collected. Instead of measuring the generator operating and environmental
quantities such as kV, MW, MVAr, hotspot temperature and humidity using a direct analog feed from
the measuring detectors in the hydrogenerator – this information is obtained from the plant computer.
This greatly reduces costs, not the least of which is associated with running analog signal wiring to the
monitoring system hardware.
Since every utility seems to be using different proprietary communications systems and protocols, the
system had to be designed to be very flexible in both physical connection options and communications
protocols. For local configuration and control, the acquisition units have an RS232 port. For remote
operation and integration with other systems, an RS485 port is utilized. An addressing scheme on the
RS485 bus allows for connection of up to 127 instruments (Figure 1).
Ac
D
qui
a
rin
t
g
a
Ne
w
Da
ta
NQ
NAl
/
ar
QM
ax
m
Hyd
roT
rac
Ca
nc
el
O
K
Att
en
tio
n
Ac
D
qui
a
rin
t
g
a
F
F
1
2
Ne
w
Da
ta
NQ
N/
Al
ar
QM
am
x
Hyd
roT
rac
Ca
nc
el
O
K
S
Com
er
muni
ia
catio
nl
Att
en
tio
n
Partial Discharge
Monitoring For
Generators
Caution. Do not remove
panel. High voltage may
be present inside.
F
F
1
2
S
Com
er
muni
ia
catio
nl
Partial Discharge
Monitoring For
Generators
Caution. Do not remove
panel. High voltage may
be present inside.
RS232 to RS485
Converter
PD System Controller
Ac
D
qui
a
rin
t
g
a
Ne
w
Da
ta
NQ
NAl/
ar
QM
am
x
Hyd
roT
rac
Ca
nc
el
O
K
Plant Ethernet LAN/WAN
Att
en
tio
n
Com
er
muni
ia
catio
nl
Partial Discharge
Monitoring For
Generators
Caution. Do not remove
panel. High voltage may
be present inside.
File Server
Ac
D
qui
a
rin
t
g
a
F
F
1
2S
Ne
w
Da
ta
NQ
NAl/
ar
QM
am
x
Ca
nc
el
O
K
Att
en
tio
n
F
F
1
2S
Hyd
roT
rac
Com
er
muni
ia
catio
nl
Partial Discharge
Monitoring For
Generators
Caution. Do not remove
panel. High voltage may
be present inside.
Remote User
Figure 1: Connection of multiple PD monitors to a single control computer. Each monitor measures the PD
from up to 12 PD sensors inst alled on a generator, separates PD from noise, and then determines the number and
magnitude of PD pulses from each sensor.
[email protected]
4
The RS485 network can physically be fiber optic or shielded twisted pair cable. Ethernet
communication is possible through any number of third-party terminal servers to convert either one or
more monitors to TCP/IP Ethernet. At that point, additional communications options such as WiFi are
possible.
Utilizing any of these physical communications schemes, the system can be controlled via a remote
Windows™ application. This control software is used to configure the monitors at each
hydrogenerator, trigger PD data collection on a defined schedule, and archive the data in a database for
later review. Even if communications should fail to the control application, the instrument continually
collects and archives data in its local memory, storing up to 24 months of summary PD data. This data
can also later be accessed either locally or remotely communications for trending and review.
The control software also has the ability to trigger PD data collection based on the generator state. PD
levels fluctuate with parameters such as MW, MVAr, stator winding temperature, and humidity.
Knowing these variables can smooth PD trends and provide additional diagnostic information on the
stator winding condition. These parameters can be fed to system via a ‘PI’ interface from a plant PI
server (OSI Soft TM ). Many other plant SCADA and diagnostic software packages also contain bridges
to PI.
Some plants prefer to use 4-20 mA analog outputs to facilitate trending the PD activity over time on
plant SCADA, DCS, or other monitoring systems. In addition to the physical interfaces described
above, external systems can connect using an ‘OPC’ software interface. In this case, the OPC Server
software must be running on a Windows™ computer connected to each continuous PD monitor. This
server exposes an OPC Data Interface to any number of OPC clients. The OPC Server controls the
PD monitors and exposes PD summary data and pulse magnitude analysis data as a set of OPC tags.
Finally, the system has an alarm output relay, which can be wired to a remote system and will be
activated when predetermined PD levels are exceeded.
PRACTICAL IMPLEMENTATION
The first monitors were first installed in May 2001. Since then monitors have been installed on over
150 hydrogenerators around the world. Examples of these are described below.
NYPA St. Lawrence – Franklin D. Roosevelt Project
This plant is located on the St Lawrence River, and consists of 16 hydrogenerators rated approximately
60 MVA and 13.8 kV. They are dispersed along a >400m long stretch of the dam. The plant was
originally commissioned in the 1958. As part of a large upgrade and modernization program, all new
control and monitoring systems have been installed. Included in the modernization is the development
and installation of a generator expert monitoring system that monitors virtually all on-line systems for
the generator and turbine [7]. Thus a PD monitoring system was required that could communicate
with the expert system.
Initially, the 16 units at the plant were outfitted with continuous PD monitors connected via two
RS485 twisted pair ‘daisy chains’, one for each of 8 units due to the length of thedam. Two terminal
servers, located centrally in the powerhouse, converted the RS485 to Ethernet and connected the PD
monitors to the plant LAN. The control software was used with time-based triggers to initiate PD data
collection on a weekly basis. As part of the plant modernization process, LAN drops are being added
at each modernized unit control board close to the location of the automated PD monitors. As each
unit is refurbished, the monitors are being removed from the RS485 daisy chain, and connected via
Ethernet to an OPC Server, which then communicates with the Expert System. Rules in the expert
system trigger PD data collection and instantaneous analysis through the OPC interface at various load
and stator temperatures.
[email protected]
5
Figure 2 shows the relat ive PD activity on all 16 units at the St. Lawrence project. Figure 3 shows the
trend in PD for one unit over a 2-year period, under the same operating conditions. NYPA is now
installing the new continuous monitoring on 13 units at Lewiston pump generating project and on 12
hydrogenerators at the Robert Moses Dam. Both plants are near Niagara Falls.
1200
1000
800
Qm +
600
Qm -
400
200
0
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Figure 2. A comparison of the peak positive and negative PD activity (vertical scale, in mV) on all 16 machines
(Units 17-32, horizontal scale) equipped with continuous PD monitoring at the NYPA St. Lawrence - Franklin
Roosevelt project. Unit 18 has the most severe insulation deterioration.
Figure 3. Trend in PD in B phase over a 2-year period recorded under the same operating condition on St.
Lawrence Unit 17. There is a some increase in B phase positive PD activity, but most of the PD sensors are
showing stable PD activity over time. This, together with the PD magnitudes in comparison to similar machines
[6], indicates the winding insulation condition is starting to be of concern.
B.C. Hydro Plants
B.C. Hydro has installed continuous monitoring units on three different machines in three plants. One
has been in place since 2003 and the most recent one since mid 2005. All units are connected to B.C.
Hydro’s PI System through an OPC interface.
[email protected]
6
The first installation was on a 233 MVA machine that was newly commissioned in 2003. The
machine is rated at 13.8 kV and has four parallel circuits per phase. The initial PD test in April 2003
(using a PDA-IV) showed high partial discharge activity in the parallel circuits of “A” phase. The
initial full load hot (FLH) test results for “A” phase coupler T1-4 showed PD up to 550 mV – which is
higher than about 95% of similar machines. Couplers T1-1 and T1-2 also had high PD.
The cause of the high PD levels was unknown. A continuous monitoring instrument was installed to
monitor the PD activity over time to ensure that it did not suddenly increase and jeopardize the reliable
operation of the generator. Periodic tests with a portable PDA-IV instrument were also carried out to
confirm the data obtained by the continuous system. Based on the discharge patterns as shown in
Figure 4 the source was suspected to be in the end turn region.
Coupler C2
10 4
POS
NEG
Pulse Count Rate
[PPS]
10 3
10 2
10 1
0
100
200
300
400
500
Pulse Magnitude [mV]
600
700
800
Figure 4. Plot of PD repetition rate (pulse per second, vertical scale) vs. PD magnitude (mV, horizontal scale)
for Circuit T1-2.
A visual inspection in October of 2003 confirmed the source to be a manufacturing defect in the bars
immediately outside the slot. This defect reduced the clearance between the front and back bars of the
stator winding. These reduced clearances combined with the higher voltage between front and back
bars when they were in different phases, resulted in partial discharge activity between some front and
back bars.
As shown in the trend data in Figure 5, the PD magnitude in circuit T1-4 decreased slightly over time.
One reason for the decrease was cleaning of the upper end turns in November 2003. A subsequent
improvement occurred after March 2004 when an experimental repair was carried out in two locations
in circuit T1-4. The repair consisted of filling the gaps between two front and two back bars operating
near line potential in different phases. Also note that lower PD values occurred after June 2004 and
April 2005, and are as yet unexplained. Data obtained from the PI system indicated the reductions
were not due to the operating conditions. They may have been caused by environmental factors.
Continuous monitoring of this winding is ongoing while permanent repairs are developed to eliminate
the discharge between front and back bars.
Qm trend plot
Qm C1_POS_MAX
Qm C1_NEG_MAX
1500
1250
1000
m
V 750
500
250
2003
Oct
Jan 2004
Apr
Jul
Time
Oct
Jan 2005
Apr
Jul
Figure 5. PD magnitude (in mV) vs. date over 2 years for Circuit T1-4 of a new 233 MVA winding
[email protected]
7
The second installation occurred in 2004 on a 35 year old 275 MVA, 13.8 kV stator winding. This
machine had 6 circuits per phase. After repairs in 2002 to eliminate PD from the circuit ring bus the
PD activity decreased substantially. Then in 2004 during routine PD testing a sudden increase in
activity was observed in “C” phase circuits T3A3 and T3B4 as shown in Table 1. The high PD is very
localized, and is plotted in Figure 6.
Table 1. PDA IV FLH test results
Circuit
Qm+ mV
Qm- m V
129
96
1066
2089
84
98
99
97
987
1979
75
95
T3A1
T3B2
T3A3
T3B4
T3A5
T3B6
Coupler C1
104
Coupler C2
104
POS
POS
NEG
NEG
103
Pulse Count Rate [PPS]
Pulse Count Rate [PPS]
103
102
101
102
101
0
100
200
300
400
500
Pulse Magnitude [mV]
600
700
800
0
500
1000
1500
2000
Pulse Magnitude [mV]
2500
3000
Figure 6. PD repetition rate vs. magnitude for Circuit T3A3 and T3B4
Through analysis of the PD data as a function of AC phase position, it was determined that the PD
activity was likely located in the winding end turns. PD in this area was considered a very high risk to
the reliable operation of the machine because the end turns were suspected to contain minute cracks.
A short outage was arranged and an immediate investigation was carried out to identify the location
and carry out the necessary repairs. Three active PD sites were identified in the circuits that had the
high PD measurements. These areas were cleaned and repaired with tape and resin and the unit
returned to service. Time did not permit more widespread repairs. Subsequent PD tests showed that
the PD level in the two circuits of concern were reduced to that shown in Figure 7 but the amplitude
was still unacceptably high in circuit T3B4.
Coupler C1
104
Coupler C2
104
POS
POS
NEG
NEG
103
103
Pulse Count Rate [PPS]
Pul
se
Co
unt
Rat 2
e10
[P
PS
]
101
102
101
0
500
1000
1500
2000
Pulse Magnitude [mV]
2500
3000
0
500
1000
1500
2000
Pulse Magnitude [mV]
2500
3000
Figure 7. PD activity for Circuits T3A3 and T3B4 Plots After Initial Repair
[email protected]
8
A second outage was scheduled and a more thorough attempt made to clean the winding. At this time
it was noted that the PD was not isolated to only a few locations but was a more general condition in
the upper and lower end turns between bars in different phases. Once the cleaning of the end turns
was complete, the winding was coated with an insulating resin and the machine returned to service.
The continuous on -line PD monitor was installed a few months prior to the outage to provide trending
information about the PD activity in this machine. Once again it was confirmed that the high
amplitude discharges were reduced (Figure 8). However, the PD activity on this machine continues to
be monitored and an operating policy has been implemented to minimize the hours of operation and
the number of starts and stops.
Qm trend plot
Qm trend plot
Qm C1_POS_MAX
1750
Qm C1_NEG_MAX
Qm C2_POS_MAX
Qm C2_NEG_MAX
3000
1500
2500
1250
2000
mV
mV
1000
1500
750
1000
500
500
250
Mar
Apr
May
2005
Jun
Time
Jul
Aug
Sep
Mar
2005
Apr
May
Jun
Time
Jul
Aug
Sep
Figure 8. PD magnitude (vertical scale) for circuits T3A3 and T3B4 for the period April to September 2005
(horizontal scale).
In both of the above machines the monitoring instrument has proved a valuable aid in recording the
PD activity and providing early warning of increases in PD activity that may jeopardize the reliable
operation of the machines. In addition, the PD data can be compared with other machine parameters
recorded by the PI system. The instrument has provided an inexpensive way to monitor partial
discharge activity accurately without interrupting machine operation.
CONCLUS ION
Continuous on -line PD monitoring that can be easily integrated with hydrogenerator plant computer
system is now a proven, cost-effective means to determine maintenance requirements for stator
winding insulation systems. These systems enable plant maintenance staff to identify which stators
need attention, and to evaluate the effectiveness of corrective maintenance. Utilizing such an
automated continuous system will allow links to expert system diagnostic software.
BIBLIOGRAPHY
[1]
[2]
[3]
[4]
G.C. Stone, E.A. Boulter, I. Culbert, and H. Dhirani, “Electrical Insulation for Rotating
Machines”, Wiley Interscience, 2004.
J. Johnson, M. Warren, “Detection of Slot Discharges in High Voltage Stator Windings During
Operation”, Trans AIEE, Part II, 1951, pp 1993-2000.
J.E. Timperly, E.K. Chambers, “Locating Defects in Large Rotating Machines and Associated
Systems Through EMI Diagnostics”, CIGRE Paper 11-311, September 1992
IEC 60034: “Rotating Electrical Machines – Part 27: Partial Discharge Off-line Measurements
on the Stator Winding Insulation of Rotating Machinery”, Draft, 2005.
[email protected]
9
[5]
M. Kurtz et al, “Diagnostic Testing of Generator Insulation Without a Service Interruption”,
CIGRE Paper 11-09, September 1980.
[6] G.C. Stone, et al, “Advances in Interpreting PD Test Results from Motor and Generator Stator
Windings.”, CIGRE, Paper 11-202, September 2002.
[7] M. Krikorian et al, “Fielding a Knowledge Based Diagnostic Monitoring System for HydroGenerators”, Proc. HydroVision Conference, Montreal, July 2004.
[8] E. Binder et al,
"Developments and Verification Tests of Diagnosis Methods for
Hydrogenerators", CIGRE Paper 11-301, September, 2000
[9] E. Colombo, G. Tontini, “On-line PD measurement on the Stator Winding”, Proc IEEE
International Symposium on Electrical Insulation, June 1992.
[10] B. Lloyd, S.R. Campbell, G.C. Stone, “Continuous On-Line Partial Discharge Monitoring of
Generator Stator Windings,” IEEE Transactions on Energy Conversions. Dec. 1999, pp11311137.
[11] C. Marsh et al, ‘Real Time Assessed Condition Trending Using Continuous PD Monitoring’,
Utility Generator Predictive Maintenance and Refurbishment Conference, EPRI, Jan. 2001.
[12] R. Viteri, “Experience with an Automated Continuous PD Monitoring System for
Hydrogenerators”, Iris Rotating Machine Conference, March 1998.
[email protected]
10