3 APPLICATION OF PD-CPWA
3.1 PD IN SOLID DIELECTRICS
3.1.1 EXPERIMENTAL SETUP AND PROCEDURE
Application of PD-CPWA to PD in an epoxy spacer for GIS
is introduced. Figure 7 shows the experimental setup of an
epoxy spacer model [7]. High-voltage and grounded electrodes
made of aluminum with the edge radius of 6 mm were
embedded in the epoxy resin. The gap length between both the
electrodes was 4 mm. The virgin sample of the epoxy spacer
model was PD free up to 154 kVrms under ac 60 Hz voltage
application.
Simulating the operational condition of GIS, thermal stress
was applied to the epoxy spacer models in an oven for
different temperatures and durations. Table 1 summarizes the
thermal stress conditions for total 9 samples. Each sample was
placed in SF6 gas at 0.4 MPa. Then, electric stress was applied
to the sample and generated PD. After the PD inception, the
applied voltage was kept at the PD inception voltage (PDIV).
The sequential PD generation characteristics were obtained by
PD-CPWA from PD inception to BD.
3.1.2 EXPERIMENTAL RESULTS
Figure 8 shows PD generation characteristics from PD
inception to BD for the sample with thermal stress of 180 ºC
for 5 h. PDIV was 141.0 kVrms and BD was induced at 11
minutes 22 seconds after the PD inception. Total 5,366 PD
pulses (positive: 3,340, negative: 2,026) were generated from
PD inception to BD. Such statistical data of PD generation
characteristics for each sample are shown in Table 1.
Figures 9a to 9g show the time transitions of PD parameters
analyzed by PD-CPWA for the last 60 s before BD in Figure
8: (a) peak value I with their waveforms, (b) di/dt, (c) rise time
tr, (d) fall time tf, (e) PD charge q, (f) PD energy J of each PD
current pulse waveform, respectively, and (g) accumulated PD
energy Ja. q was calculated by the integration of PD current
pulse waveform, and J was evaluated by the product of PD
charge and applied voltage. The following tendencies can be
derived from Figure 9(1) PD current pulse waveform becomes larger and steeper
with the elapse of time.
(2) PD generation was not continuous, but intermittent
from PD inception to BD.
(3) Time transitions of di/dt, q and J were almost consistent
with that of peak value I.
(4) Time transitions of tr and tf were not consistent with
that of peak value I.
(5) Polarity difference in each PD parameter was not
significant.
(6) Accumulated PD energy Ja gradually increased and
reached about 240 μJ, in this sample, just before BD.
3.1.3 DISCUSSIONS
The above PD generation characteristics have a certain
tendency to BD in most samples, irrespective of the difference
in thermal stress conditions. Figures 10 and 11 show the time
transition of I and di/dt for 8 samples resulted in BD within 3
hours, respectively, where PDIV, time to BD and number of
PD pulses for each sample are listed in Table 1. Figure 12
shows the time transition of time interval Δt of subsequent PD
pulses for 4 samples. Note that the horizontal time axes in Figs.
10, 11 and 12 are normalized by the time to BD after PD
inception in each sample. The drastic increase in I and di/dt
before BD was also consistent with the decrease in Δt into the
order of microseconds. Such time transition of PD parameters may be attributed to
the change of physical properties of the epoxy resin under the
thermal and electric field stresses. The insulation performance
of epoxy resin would be deteriorated due to electron impact,
local temperature rise, chemical reaction and so on at the
interface between the epoxy resin and the embedded electrode.
The insulation degradation of the epoxy resin could further
activate the PD development in the epoxy resin. Such a
positive feedback process between the PD generation and the
insulation degradation of epoxy resin would be accumulated
during the long time voltage application and result in BD when
a specific PD parameter reached the critical level.
PD-CPWA is expected to contribute to identify the specific
PD parameter to be closely related to the physical mechanisms
of insulation degradation leading to BD in epoxy spacer for
GIS. PD-CPWA can also be applied to the PD measurement
and analysis for XLPE cables.
3.2 CREEPAGE PD IN SF6 GAS
3.2.1EXPERIMENTAL SETUP AND PROCEDURE
Figure 13 shows the experimental setup for creepage PD
measurement in SF6 gas [8]. A needle was fixed on an epoxy
plate between parallel plane electrodes in SF6 gas at 0.1 MPa
(dimensions are shown in Figure 13). In order to measure PD
current pulse waveform generating at the needle tip under ac
60 Hz voltage application, a detecting and matching circuit
with 50 Ω resistor was set under the epoxy plate. The detected
PD signals were fed into the digital oscilloscope and analyzed
by PD-CPWA. In addition, PD light emission image was also
observed by a digital camera thorough an image intensifier.
Table 2 shows different arrangements of needle and epoxy
plate with the measured PDIV. When PD-CPWA was applied
to each arrangement, the applied voltage was set at about 160
% of PDIV, because the number of PD signals to be shown
later decreased with the elapse of time. PD measurement under
PD-CPWA was repeated in every 2 minutes for 30 minutes.
3.2.2 EXPERIMENTAL RESULTS AND
DISCUSSIONS
Figure 14 shows (i) typical PD current pulse waveforms in
the negative cycle and (ii) PD light emission images for 1 s,
respectively, at (a) t = 0 min and (b) t = 30 min after the
voltage application, when the needle was fixed on the middle
of epoxy plate. The peak value of PD current pulse waveform
at t = 30 min decreased drastically into 1/20 of that at t = 0 min.
Such a relaxation of PD activity can also be verified by the PD
light emission images, where the PD light intensity at both
ends of the needle at t = 0 min was much higher than that at t =
30 min. On the other hand, the rise time and fall time in the PD
current pulse waveform at t=0 min were nearly equal to those
at t = 30 min, respectively, which means that di/dt decreased
with the elapse of time.
Figure 15 shows the time transition of averaged values and
standard deviations of positive and negative PD current pulses
in every 2 minutes, when the needle was fixed on the middle of
epoxy plate. The averaged values of PD current pulses
decreased drastically within the first 2 minutes and then
gradually decreased or almost constant after 2 minutes. Figure
16 shows the time transition of di/dt in the negative cycle for
different arrangements in Table 2. Note that di/dt in Figure 16
are normalized by those just after the voltage application. Thetime transition of di/dt depended on the location of the needle
fixed on the epoxy plate. The time constants of the decrease in
di/dt were about 1, 2 and 5 minutes, respectively. The
difference in the time constant will be interpreted by the
charging characteristics and mechanisms on the epoxy plate as
well as the resultant distortion of electric field distribution for
each arrangement.
The above measurement and discussion on the transition of
PD parameters using PD-CPWA can contribute to locate the
fixed particles on epoxy spacer and distinguish them from free
particles in GIS.
3.3 PD IN LIQUID/SOLID COMPOSITE SYSTEM
3.3.1 EXPERIMENTAL SETUP AND PROCEDURE
Another application of PD-CPWA to PD in liquid/solid
composite system is introduced. Figure 17 shows the
experimental setup of liquid nitrogen / polypropylene
laminated paper composite insulation system for high
temperature superconducting cables [9]. A coaxial cylindrical
cable model with the insulation thickness of 1 mm (0.125 mm
x 8 layers) and the effective length of 150 mm was immersed
in liquid nitrogen at 77 K. In the similar way as those in the
previous sub-sections, PD current pulse waveform under ac 60
Hz voltage application was measured through 50 Ω resistor
and analyzed by PD-CPWA.
3.3.2 EXPERIMENTAL RESULTS AND
DISCUSSIONS
PDIV of the cable model was 18 kVrms and BD was induced
at 35 kVrms. Figure 18 shows the time transitions of peak value
of PD current pulse waveform at the applied voltage of (a) 20
kVrms, (b) 26 kVrms and (c) 35 kVrms, respectively. The peak
value and number of PD pulses per second increased with the applied voltage. PD pulses at the lower applied voltage were
generated at around the peak of the applied voltage, whereas
those at the higher applied voltage also appeared at around the
zero-crossings. These results suggested that the PD generation
mechanisms changed into void-like discharges not only in butt
gaps but also in the other micro gaps between the laminated
paper layers [9].
Figure 19 shows typical PD current pulse waveforms at the
applied voltage of (a) 20 kVrms and (b) 27 kVrms, respectively.
The averaged values of rise time and fall time in PD current
pulse waveform were about 18 ns and di/dt was about 0.02
mA/ns, irrespective of the applied voltage. These values were
longer and slower than those of the epoxy spacer and creepage
PD in SF6 gas in the previous sub-sections. Even the
outstanding large PD pulses in a cluster of small PD pulses in
Figure 18c had the similar rise time, i.e. steeper di/dt,
compared with those in the cluster. Such a contamination of
large PD pulses suggests the superposition of PD pulses with
different PD mechanisms from that of void-like discharges. By
analyzing the PD parameters, we will be able to identify the
PD location in butt gaps or laminated paper layers and liquid
or gas phase, etc.
As described above, PD-CPWA was verified to contribute
to discuss the change of PD generation characteristics and
their physical mechanisms for different electrical insulating
materials and circumstances.
แอพลิเคชัน 3 ของ PD-CPWA3.1 PD ใน DIELECTRICS แข็ง3.1.1 ทดลองตั้งค่าและขั้นตอนโปรแกรมประยุกต์ของ PD CPWA PD ในเป็นตัวเว้นวรรคอีพ๊อกซี่สำหรับ GISมีการแนะนำ รูปที่ 7 แสดงการทดลองการอีพ๊อกซี่เป็นตัวเว้นวรรครุ่น [7] แรงดัน ไฟฟ้าสูง และป่นเล็กน้อยหุงตทำจากอลูมิเนียม มีรัศมีขอบ 6 มม.ได้ฝังอยู่ในยางเรซินสังเคราะห์ ความยาวของช่องว่างระหว่างทั้งสองหุงต 4 มม.ได้ ตัวอย่างเป็นตัวเว้นวรรคอีพ็อกซี่บริสุทธิ์รุ่นฟรีถึง 154 kVrms ภายใต้แรงดัน ac 60 Hz PDแอพลิเคชันการจำลองสภาพการดำเนินงานของ GIS ความเครียดความร้อนมีรูปเป็นตัวเว้นวรรคแบบอีพ็อกซี่ในเตาอบสำหรับใช้อุณหภูมิแตกต่างกันและระยะเวลา ตารางที่ 1 สรุปการสภาวะความเครียดความร้อนตัวอย่างรวม 9 แต่ละตัวอย่างถูกวางในก๊าซ SF6 ที่ 0.4 แรง ใช้แล้ว ความเครียดไฟฟ้าการปอนด์ชิ้นงานตัวอย่าง และสร้าง หลังจากมา PD การใช้แรงดันไฟฟ้าถูกเก็บไว้ที่แรงดันไฟฟ้ามา PD (PDIV)ลักษณะรุ่น PD ตามลำดับได้รับโดยPD-CPWA จาก PD มากับ BD3.1.2 ผลการทดลองรูปที่ 8 แสดงลักษณะรุ่น PD จาก PDมากับ BD สำหรับตัวอย่างกับความเครียดความร้อนของ 180 ºC5 h. PDIV ได้ 141.0 kVrms และ BD มีเกิดที่ 11นาที 22 วินาทีหลังจากเริ่ม PD รวม 5,366 PDกะพริบ (บวก: ลบ 3,340:2,026) สร้างขึ้นจากมา PD กับ BD เช่นข้อมูลทางสถิติของรุ่น PDลักษณะสำหรับแต่ละตัวอย่างจะแสดงในตารางที่ 1ช่วงเวลาของ PD พารามิเตอร์แสดงเลข 9a 9 กรัมวิเคราะห์ โดย PD CPWA 60 สุดท้าย s ก่อน BD ในรูป8: (ก) ยอดมูลค่าของ waveforms, (b) di/dt เวลา (c) เพิ่มขึ้นtr รหัสเวลาฤดูใบไม้ร่วง (d), (e) PD ค่า q, J พลังงาน (f) PD ของ PD แต่ละปัจจุบันหมุนรูปคลื่น ตามลำดับ และ PD สะสม (กรัม)พลังงานที่จะ q คำนวณ โดยรวมของ PD ปัจจุบันหมุนรูปคลื่น และ J ถูกประเมิน โดยผลิตภัณฑ์ของ PDใช้แรงดันไฟฟ้าและค่า แนวโน้มต่อไปนี้ได้มาจาก 9(1) รูป PD รูปคลื่นพัลส์ที่ปัจจุบันกลายเป็นขนาดใหญ่ และสูงชันมีพ้นเวลา(2) PD สร้างไม่ต่อเนื่อง แต่ไม่ต่อเนื่องจาก PD มากับ BD(3) ช่วงการเปลี่ยนภาพเวลา di/dt, q และ J ได้เกือบตรงมีค่าสูงสุดของฉัน(4) เวลาเปลี่ยน tr และ tf ไม่สอดคล้องกับของสูงค่าฉัน(5) ไม่มีขั้วความแตกต่างในแต่ละพารามิเตอร์ PDอย่างมีนัยสำคัญ(6) สะสมพลังงาน PD จะค่อย ๆ เพิ่มขึ้น และถึงประมาณ 240 μJ ในตัวอย่างนี้ ก่อน BDเป็น 3.1.3 สนทนาสร้างลักษณะข้างต้นของ PD มีความบางแนวโน้มการ BD ในตัวอย่างส่วนใหญ่ โดยไม่คำนึงถึงความแตกต่างในสภาวะความเครียดความร้อน ตัวเลข 10 และ 11 แสดงเวลาเปลี่ยนแปลงของฉันและ di/dt สำหรับตัวอย่างที่ 8 ให้ BD ภายใน 3ชั่วโมง ตาม ลำดับ ที่ PDIV, BD และจำนวนเวลาPD กะพริบสำหรับแต่ละตัวอย่างแสดงในตารางที่ 1 รูปที่ 12แสดงการเปลี่ยนแปลงเวลาของ Δt ช่วงเวลาของ PD ตามมาpulses for 4 samples. Note that the horizontal time axes in Figs.10, 11 and 12 are normalized by the time to BD after PDinception in each sample. The drastic increase in I and di/dtbefore BD was also consistent with the decrease in Δt into theorder of microseconds. Such time transition of PD parameters may be attributed tothe change of physical properties of the epoxy resin under thethermal and electric field stresses. The insulation performanceof epoxy resin would be deteriorated due to electron impact,local temperature rise, chemical reaction and so on at theinterface between the epoxy resin and the embedded electrode.The insulation degradation of the epoxy resin could furtheractivate the PD development in the epoxy resin. Such apositive feedback process between the PD generation and theinsulation degradation of epoxy resin would be accumulatedduring the long time voltage application and result in BD whena specific PD parameter reached the critical level.PD-CPWA is expected to contribute to identify the specificPD parameter to be closely related to the physical mechanismsof insulation degradation leading to BD in epoxy spacer forGIS. PD-CPWA can also be applied to the PD measurementand analysis for XLPE cables.3.2 CREEPAGE PD IN SF6 GAS3.2.1EXPERIMENTAL SETUP AND PROCEDUREFigure 13 shows the experimental setup for creepage PDmeasurement in SF6 gas [8]. A needle was fixed on an epoxyplate between parallel plane electrodes in SF6 gas at 0.1 MPa(dimensions are shown in Figure 13). In order to measure PDcurrent pulse waveform generating at the needle tip under ac60 Hz voltage application, a detecting and matching circuitwith 50 Ω resistor was set under the epoxy plate. The detectedPD signals were fed into the digital oscilloscope and analyzedby PD-CPWA. In addition, PD light emission image was alsoobserved by a digital camera thorough an image intensifier.Table 2 shows different arrangements of needle and epoxyplate with the measured PDIV. When PD-CPWA was appliedto each arrangement, the applied voltage was set at about 160% of PDIV, because the number of PD signals to be shownlater decreased with the elapse of time. PD measurement underPD-CPWA was repeated in every 2 minutes for 30 minutes.3.2.2 EXPERIMENTAL RESULTS ANDDISCUSSIONSFigure 14 shows (i) typical PD current pulse waveforms inthe negative cycle and (ii) PD light emission images for 1 s,respectively, at (a) t = 0 min and (b) t = 30 min after thevoltage application, when the needle was fixed on the middleof epoxy plate. The peak value of PD current pulse waveformat t = 30 min decreased drastically into 1/20 of that at t = 0 min.Such a relaxation of PD activity can also be verified by the PDlight emission images, where the PD light intensity at bothends of the needle at t = 0 min was much higher than that at t =30 min. On the other hand, the rise time and fall time in the PD
current pulse waveform at t=0 min were nearly equal to those
at t = 30 min, respectively, which means that di/dt decreased
with the elapse of time.
Figure 15 shows the time transition of averaged values and
standard deviations of positive and negative PD current pulses
in every 2 minutes, when the needle was fixed on the middle of
epoxy plate. The averaged values of PD current pulses
decreased drastically within the first 2 minutes and then
gradually decreased or almost constant after 2 minutes. Figure
16 shows the time transition of di/dt in the negative cycle for
different arrangements in Table 2. Note that di/dt in Figure 16
are normalized by those just after the voltage application. Thetime transition of di/dt depended on the location of the needle
fixed on the epoxy plate. The time constants of the decrease in
di/dt were about 1, 2 and 5 minutes, respectively. The
difference in the time constant will be interpreted by the
charging characteristics and mechanisms on the epoxy plate as
well as the resultant distortion of electric field distribution for
each arrangement.
The above measurement and discussion on the transition of
PD parameters using PD-CPWA can contribute to locate the
fixed particles on epoxy spacer and distinguish them from free
particles in GIS.
3.3 PD IN LIQUID/SOLID COMPOSITE SYSTEM
3.3.1 EXPERIMENTAL SETUP AND PROCEDURE
Another application of PD-CPWA to PD in liquid/solid
composite system is introduced. Figure 17 shows the
experimental setup of liquid nitrogen / polypropylene
laminated paper composite insulation system for high
temperature superconducting cables [9]. A coaxial cylindrical
cable model with the insulation thickness of 1 mm (0.125 mm
x 8 layers) and the effective length of 150 mm was immersed
in liquid nitrogen at 77 K. In the similar way as those in the
previous sub-sections, PD current pulse waveform under ac 60
Hz voltage application was measured through 50 Ω resistor
and analyzed by PD-CPWA.
3.3.2 EXPERIMENTAL RESULTS AND
DISCUSSIONS
PDIV of the cable model was 18 kVrms and BD was induced
at 35 kVrms. Figure 18 shows the time transitions of peak value
of PD current pulse waveform at the applied voltage of (a) 20
kVrms, (b) 26 kVrms and (c) 35 kVrms, respectively. The peak
value and number of PD pulses per second increased with the applied voltage. PD pulses at the lower applied voltage were
generated at around the peak of the applied voltage, whereas
those at the higher applied voltage also appeared at around the
zero-crossings. These results suggested that the PD generation
mechanisms changed into void-like discharges not only in butt
gaps but also in the other micro gaps between the laminated
paper layers [9].
Figure 19 shows typical PD current pulse waveforms at the
applied voltage of (a) 20 kVrms and (b) 27 kVrms, respectively.
The averaged values of rise time and fall time in PD current
pulse waveform were about 18 ns and di/dt was about 0.02
mA/ns, irrespective of the applied voltage. These values were
longer and slower than those of the epoxy spacer and creepage
PD in SF6 gas in the previous sub-sections. Even the
outstanding large PD pulses in a cluster of small PD pulses in
Figure 18c had the similar rise time, i.e. steeper di/dt,
compared with those in the cluster. Such a contamination of
large PD pulses suggests the superposition of PD pulses with
different PD mechanisms from that of void-like discharges. By
analyzing the PD parameters, we will be able to identify the
PD location in butt gaps or laminated paper layers and liquid
or gas phase, etc.
As described above, PD-CPWA was verified to contribute
to discuss the change of PD generation characteristics and
their physical mechanisms for different electrical insulating
materials and circumstances.
การแปล กรุณารอสักครู่..
