PD are a partial breakdown of gas inclusions in insulation where the electric field intensity exceeds the breakdown field strength PD transforms part of the capacitive stored energy into heat and radiation as well as mechanical and chemical energies which can degrade insulation materials Partial Discharge Measurements energies which can degrade insulation materials This ageing process progressively reduces the insulation thickness and the breakdown voltage until the failure occurs Insulation erosion is very fast in organic materials (Type 1) slow in organic/inorganic insulation systems (Type 2)
109
Embed
Partial Discharge Measurements - Freerichard.grisel.free.fr/ICEM2012/TUTORIALS/TUT4-Part2.pdf · Partial Discharge Measurements. Electric signals related to currents, electromagnetic
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
PD are a partial breakdown of gas inclusions in insulationwhere the electric field intensity exceeds the breakdownfield strengthPD transforms part of the capacitive stored energy intoheat and radiation as well as mechanical and chemicalenergies which can degrade insulation materialsThis ageing process progressively reduces the insulationthickness and the breakdown voltage until the failureoccursInsulation erosion is very fast in organic materials (Type 1) slow in organic/inorganic insulation systems (Type 2)
Partial Discharge Measurements
PD are a partial breakdown of gas inclusions in insulationwhere the electric field intensity exceeds the breakdownfield strengthPD transforms part of the capacitive stored energy intoheat and radiation as well as mechanical and chemicalenergies which can degrade insulation materialsThis ageing process progressively reduces the insulationthickness and the breakdown voltage until the failureoccursInsulation erosion is very fast in organic materials (Type 1) slow in organic/inorganic insulation systems (Type 2)
Electric signals related to currents, electromagneticwaves, electroluminescence are recorded using propercouplers and processed in order to:
provide information on the discharge phenomenon establish PDIV as a function of the supply square waveparameters (mainly frequency and over-voltages) check the quality of the insulation and comparedifferent materials (PDIV and time-to-breakdown are currentlyadopted as parameters to be monitored during life tests)
Ref.- IEC 61934TS: Electrical insulating materials andsystems – Electrical measurement of partial discharges(PD) under short rise time and repetitive voltageimpulses
Electric signals related to currents, electromagneticwaves, electroluminescence are recorded using propercouplers and processed in order to:
provide information on the discharge phenomenon establish PDIV as a function of the supply square waveparameters (mainly frequency and over-voltages) check the quality of the insulation and comparedifferent materials (PDIV and time-to-breakdown are currentlyadopted as parameters to be monitored during life tests)
Ref.- IEC 61934TS: Electrical insulating materials andsystems – Electrical measurement of partial discharges(PD) under short rise time and repetitive voltageimpulses
PD Basics
The local breakdowngenerates a voltage dropand a consequent fastimpulsive currentabsorption from the supply
A suitable test circuit configuration allows to record the PDpulse signals. It is composed byA suitable test circuit configuration allows to record the PDpulse signals. It is composed by
Sinusoidal/square wave generator
Couplers and filters Synchro Units Digital recorders Processing unit
Very large bandwidth oscilloscopes (up to 2GHz) Quite difficult to interpret the PD measurement results Data must be organized to synthesize the information
PD usingAntenna
VoltagePulse
PD usingVoltagedivider
PD usingAntenna
ResidualNoise
Commercial instruments areavailableMust of them requiresuitable input filters to tunetheir input requirements torecord PD and reject voltagecommutations (>200-500MHz)
PD usingAntenna
VoltagePulse
PD usingVoltagedivider
PD usingAntenna
ResidualNoise
Commercial instruments areavailableMust of them requiresuitable input filters to tunetheir input requirements torecord PD and reject voltagecommutations (>200-500MHz)
v
The amplitude and the phase of the PD signal areevaluated
The PD pulse sequence is transformed in a sequence ofDirach functions having the same phase
v
PD Patterns
A 3D histogram is obtainedconsidering the number ofdischarges having the sameamplitude and the same phaseof occurrence
Two sub-patterns close tothe zero crossingAt least 1 PD per pulsePossible PD even duringthe “flat zone” of the voltagepulse
PDIV
Reference sin waveReference sin wave square wavesquare wave
PWM
VOLTAGE SUPPLY EUT
Voltagedivider
A low pass filter is required to obtain the phase referenceand to derive the so called “Phase Resolved” PD pattern
DETECTION UNIT
SYNC ChLP FILTER
Voltagedivider
Modulating wave
Modulating wave 50Hz
• Period of 20ms
Pulse Rep.Rate 1kHz
• 40 commutationsper period
Example of PRPD pattern when a PWM voltage is adopted
40 sub-patterns perperiod
Life-test Results
Weibull Plot:
)exp(1)( FttF
NF kVt
twisted pair specimens were fed till failure by differentwave forms at several amplitudes, in the presence of PDData were processed according to the standard procedure
PRPD patterns were recorded during the experiment
IPL model:
)exp(1)( FttF
Result 1: Unipolar-Bipolar Square Waves
Twisted pair of having insulation of different materialswere tested in air (with PD) and immersed in oil (no PD)
Sinusoidal, unipolar and bipolar square waves, the latterhaving the same RT, RR (50 Hz, 10 kHz), DT=50% anddifferent amplitudes were used in the test
Let V0p and Vpp the 0-to-peak and peak-to-peak values
When V0p = Vpp with or without PD, the life curves relevant tounipolar and bipolar pulses are completely overlapped
NO PD PD
The Jump Voltage is the real voltage stress that affects theinsulation ageingVEC can be adopted to compare the performances ofdifferent materials even in the presence of PDD.Fabiani et al. “Ageing acceleration of insulating materials forelectrical machine windings supplied by PWM in the presence and inthe absence of partial discharges”, IEEE ICSD pp.283-286, 2001
The picture is much more complicated when PWM andPWM+over-voltages are adopted to stress the materials in thepresence of PDDue to the “equivalent derivative effect”, there aredifferent Jump VoltagesV
VppVaa
Result 2: PWM wit and without Overvoltages
Vpp = 2(V0p +x)
x = overvoltage
V0p = voltageamplitude
Vaa = V0p +2x
t
VppVaax
F.Guastavino et al. “Life Tests on Twisted Pairs Subjected to PWM-likeVoltages”, IEEE ICSD pp.238-241, 2004
Sinusoidal
Result 3: Phase Resolved PD patterns
V = 1.83 kV V = 2.44 kV V = 3.05 kV V = 3.80 kV
Square
V = 1.9 kVV = 1.60 kV V = 4.32 kVV = 3.96 kV
Square
PWM
PWM+Over-voltage
V = 1.63 kV V = 2.73 kV V= 3.53 kV V= 4.32 kV
PWM+Over-voltage
V = 2.08 kV V = 2.50 kVV = 2.38 kVV = 2.90 kV
Discussion
PWM voltage waveforms can promote PD activity,reducing the life of magnet wires
The predominant factors explaining this behavior arepeak-to-peak voltage and switching frequency.
Even in the absence of PD, PWM voltage waveformscan accelerate the intrinsic aging of winding insulation(overvoltages). Very high slew rate (>5 kV/s), infact, plays also a non negligible role on accelerationdegradation due to increased voltage stress andheating.
This effect can be evaluate resorting to space chargemeasurements both in the presence and in the absenceof PD
Discussion
PWM voltage waveforms can promote PD activity,reducing the life of magnet wires
The predominant factors explaining this behavior arepeak-to-peak voltage and switching frequency.
Even in the absence of PD, PWM voltage waveformscan accelerate the intrinsic aging of winding insulation(overvoltages). Very high slew rate (>5 kV/s), infact, plays also a non negligible role on accelerationdegradation due to increased voltage stress andheating.
This effect can be evaluate resorting to space chargemeasurements both in the presence and in the absenceof PD
Space Charge and PD Activity
Space Charge = charge trapped inside the insulationor on interfaces (depends on the material, the poling field,the temperature and the supply voltage waveform)
Above 10 Hz, SC is not trapped appreciably in insulationunless the waveform contains a DC component
IN HF pulse waveforms, SP is accumulated mainly by PD
Neglecting the SP accumulated in the bulk, the electricfield in air, at the insulation surface, E*
0 can be:
Where:E0 : electric field without accumulated charge
0 : the permittivity of air andinsulation
d l0 : the insulation thickness and the half air gap
0: the surface charge density
Space Charge = charge trapped inside the insulationor on interfaces (depends on the material, the poling field,the temperature and the supply voltage waveform)
Above 10 Hz, SC is not trapped appreciably in insulationunless the waveform contains a DC component
IN HF pulse waveforms, SP is accumulated mainly by PD
Neglecting the SP accumulated in the bulk, the electricfield in air, at the insulation surface, E*
0 can be:
Where:E0 : electric field without accumulated charge
0 : the permittivity of air andinsulation
d l0 : the insulation thickness and the half air gap
0: the surface charge density
0000
*
ld
dEE s
Behaviour of the electric field in two enamelled wires in atwisted pair configuration
This configuration helps explain why the main ageingfactor is associated to the jump voltage for bipolar andunipolar voltage waveforms
Let E*0(t) the behaviour of the electric field in the air gap
before and after a PD event (gray line)
E0(t) the applied field (black line)
-E0p +E0p the bipolar voltage amplitude
2E0p the unipolar voltage amplitude
and let E0>PDIVESC the drop of the electric field due to the PD chargeinjection
Let E*0(t) the behaviour of the electric field in the air gap
before and after a PD event (gray line)
E0(t) the applied field (black line)
-E0p +E0p the bipolar voltage amplitude
2E0p the unipolar voltage amplitude
and let E0>PDIVESC the drop of the electric field due to the PD chargeinjection
After PD, a residualvalue of the electricfield is:
00
0*
0* 0
ld
dE
EEE
s
scres
If the chargeinjected by PD isnot rapidlydepleted, E*
The influence of the JV was experimentally supplying 4different kinds of enamelled insulated wires with sinusoidal,unipolar and bipolar square waves in the range of 50 to 10kHz and with a RT of 50 ns above the PDIV
SC measurement becomes a significant tool to evaluate theability of the different materials to deplete the SC injectedby PD and to compare different enamels
PEA SP Measurement System
A Pulse Electro-Acustic method ha been developed tomeasure the space charge accumulation on magnet wires
(D. Fabiani et al. “ Relation Between Space Charge Accumulation andPartial Discharge Activity in Enameled Wires Under PWM-like VoltageWaveforms”, IEEE Trans. on Diel.-Elect.Ins., Vol.11, pp.393-405, June2004)
The specimen is positioned analuminium ground plate and
a semicon-absorber and
fed by HV DC power supply
A voltage pulse of A=300 V, 10ns width, 110Hz of RR, is appliedthrough a 220 pF couplingcapacitor
The specimen is positioned analuminium ground plate and
a semicon-absorber and
fed by HV DC power supply
A voltage pulse of A=300 V, 10ns width, 110Hz of RR, is appliedthrough a 220 pF couplingcapacitor
220 pF
Pulse
2 M
HVDC
Adsorber (PMMA)
Piezoelectric (PVDF)
Semicon AdsorberEnameled Insulation
Copper
Oscilloscope
GPIB IEEE-488Acquisition Board
PC
Amplifier
PVC Insulation
Aluminum Ground Plate 220 pF
Pulse
2 M
HVDC
Adsorber (PMMA)
Piezoelectric (PVDF)
Semicon AdsorberEnameled Insulation
Copper
Oscilloscope
GPIB IEEE-488Acquisition Board
PC
Amplifier
PVC Insulation
Aluminum Ground Plate
Charges present in insulation are forced by the fastelectric pulse and a pressure wave is generated from theinteraction between charges and the material structure
The pressure wave propagates through the insulation andreaches the ground electrode under which the piezoelectrictrasducer (PVDF) is located
PVDF generates a voltage signal (PEA output signal)proportional to the pressure wave propagating through it
A proper calibration procedure allows the PEA outputsignal to be correlated to the amount of trapped spacecharge
PEA signal is amplified and sent to a recording system
110 PEA outputs per second, synchronized with therectangular supply voltage, are processed to obtain thecharge profile
Charges present in insulation are forced by the fastelectric pulse and a pressure wave is generated from theinteraction between charges and the material structure
The pressure wave propagates through the insulation andreaches the ground electrode under which the piezoelectrictrasducer (PVDF) is located
PVDF generates a voltage signal (PEA output signal)proportional to the pressure wave propagating through it
A proper calibration procedure allows the PEA outputsignal to be correlated to the amount of trapped spacecharge
PEA signal is amplified and sent to a recording system
110 PEA outputs per second, synchronized with therectangular supply voltage, are processed to obtain thecharge profile
Test Procedure
SC measurements are performed according to a specificpolarization/depolarization procedure
Polarization: the electric field is applied (volt-on, VP=1 kV) fora period tp=3600 s to achieve steady state conditions forthe accumulated charge
Depolarization: depolarization(volt-off) follows polarization. Itis obtained removing the supplyvoltage and grounding the highvoltage electrode, and lasted3600 s as well.
Depolarization: depolarization(volt-off) follows polarization. Itis obtained removing the supplyvoltage and grounding the highvoltage electrode, and lasted3600 s as well.
t =20 s:
PEA signal is mainly due to the electrode field-inducedcharge (peaks indicate the electrode location, i.e., anode and cathode,corresponding to the positive and negative signal peaks, respectively)
the injected charge at the electrode-insulation interface ishidden by the electrode charge
t =3602 s: SC in the insulation bulk canbe observed looking at the PEAprofile under volt-off (shadedarea)
t =3602 s: SC in the insulation bulk canbe observed looking at the PEAprofile under volt-off (shadedarea)
when the poling field is removed the electrode charge isconsiderably smaller than under polarization, being onlydue to the image charge (of the internal charge)
Space charge profile at the beginning ofvolt on: no space charge present
Space charge profile 2s after voltageremoval: space charge = gray area
To quantify the charge accumulation:
Total absolute stored charge density(after grounding), QM
Depolarization characteristic
Space Charge Data Processing
1
0
),(1
)(01
x
x
VO dxtxQxx
tQ
Total absolute stored charge density(after grounding), QM
TIPO A - PROFILO CARICA: Volt-ON 3600 s (+) Volt-OFF 3600 s
Time [s]
1 10 100 1000 10000
Cha
rge
[p.u
.]
0.2
0.4
0.6
0.8
1.0
1.2
Depolarization characteristic
the slope of depolarization characteristic,s, is a measure of space charge dynamic,i.e. the speed of charge recombination /
expulsion
Material Improvements
PD when active, are the most important ageing factor inType I insulation
Even in the absence of PD, jump voltage and switchingfrequency can accelerate the intrinsic aging of windinginsulation due to increased voltage stress and heating
Solutions could come from the use of:
VPI technologies
mica-films for turn and strand insulation
with metal oxide or ceramic fillers (micro-fillers)
nano-scale technologies (nano-fillers)
PD when active, are the most important ageing factor inType I insulation
Even in the absence of PD, jump voltage and switchingfrequency can accelerate the intrinsic aging of windinginsulation due to increased voltage stress and heating
Solutions could come from the use of:
VPI technologies
mica-films for turn and strand insulation
with metal oxide or ceramic fillers (micro-fillers)
nano-scale technologies (nano-fillers)
VPI Technologies
The stator can be totally impregnated adopting the VPItechnology even for small size random wound machines
The PD inception voltage is up to 60% higher comparedwith non-impregnated ones because air gaps are filled(not totally) by the impregnation
The time to breakdown of the inter-turn insulationdepends on the: PDIV and intensity of PD enamel thickness and its resistance against PD erosion
But due to small imperfections, longer lifetime is notguaranteed VPI process for small machines is expensive
The stator can be totally impregnated adopting the VPItechnology even for small size random wound machines
The PD inception voltage is up to 60% higher comparedwith non-impregnated ones because air gaps are filled(not totally) by the impregnation
The time to breakdown of the inter-turn insulationdepends on the: PDIV and intensity of PD enamel thickness and its resistance against PD erosion
But due to small imperfections, longer lifetime is notguaranteed VPI process for small machines is expensive
Mica-Films for Inter-Turn or Strand Insulation
They are also adopted in LVmachines
Improvements were found but
higher costs
slot efficiency reduction
increased dimensions
Currently adopted in MV and HVmachines fed with pulsating voltages
They are also adopted in LVmachines
Improvements were found but
higher costs
slot efficiency reduction
increased dimensions
Micro Fillers
Enamel or polymer film may also contain additives suchmetal oxides or ceramic materials with a naturalresistance to discharges (Al2O3, TiO2, SiO2, Fe2O3..), hasbeen adopted
The structure is a multi-coatingover the copper enameled wirewith the PD shield layer and thesurface protection coating
The structure is a multi-coatingover the copper enameled wirewith the PD shield layer and thesurface protection coating
The basic idea is to substitute the largemica flakes with a high densityinorganic materials that
form barriers for the tree growth
facilitate the space charge diffusion
facilitate the heat transmission
the endurance life is enhanced no variations in the ground-wall insulation thickness increased t2t PD and surge withstand capabilityButIt become brittle and develop cracks when subjected totemperature variation in the presence of mechanicalstressesEach micro-composite material mustbe evaluated carefully
the endurance life is enhanced no variations in the ground-wall insulation thickness increased t2t PD and surge withstand capabilityButIt become brittle and develop cracks when subjected totemperature variation in the presence of mechanicalstressesEach micro-composite material mustbe evaluated carefully
Tests:Tests: -- in the absence of PD (in oil), below PDIVin the absence of PD (in oil), below PDIV-- in the presence of PD (in air), above PDIVin the presence of PD (in air), above PDIV
Sinusoidal and distorted waveformsSinusoidal and distorted waveforms
Evaluation of Corona Resistant Materials
Objective: to draw life linesto draw life lines
life modelslife models
voltage endurance coefficient evaluationvoltage endurance coefficient evaluation
comparison among insulation systemscomparison among insulation systems
•Peak-to-peakvoltage is the mainstressing factor,besides pulserepetitionfrequency
•Effect of voltageshape negligible(experimentalpoints fit the sameline)
Tipo A - B valori di QM
Voltage frequency
DC 0.1Hz 50Hz 10kHz
QM
[C/m
3 ]
0
2
4
6
8
10
12
#A#B
Material s [s-1]#A 2.5#B 1.8
Speed of chargeexpulsion, s, for 50 Hzsquarewave voltage
(1000 V peak)
Space Charge Meas. On #A and #B
Tipo A - B valori di QM
Voltage frequency
DC 0.1Hz 50Hz 10kHz
QM
[C/m
3 ]
0
2
4
6
8
10
12
#A#B
Material s [s-1]#A 2.5#B 1.8
Total absolute stored charge density,QM, as a function of frequency. Bipolarsquarewave. Poling voltage: 1000 V
peak
Note that:QM decreases as the
frequency increasesQM (#B) > QM (#A) s (#B) < s (#A)
The CR solutions show different performance andbehaviorCR micro-composites must be evaluated carefully beforethey useD.Fabiani et al.”The Effect of Fast repetitive Pulses on theDegradation of Turn Insulation of Induction Motors”, Proc.of SDEMPED 2001, pp.289-293, Grado (I), 2001
Pulsating voltage accelerates degradation both inair and in oil (i.e. with and without PD), but differentVEC
CR materials show a longer life at higherfrequencies
The standard insulation, #A, seems to suffersignificantly from PD activity and frequency increase
CR material, #B, withstands PD better than #A
PD increase due to PWM voltage waveforms
#B tends to accumulate much more charge than#A (>QM and <s) at frequency up to 50 Hz
#B is worse than #A at 50 Hz
Discussion
Pulsating voltage accelerates degradation both inair and in oil (i.e. with and without PD), but differentVEC
CR materials show a longer life at higherfrequencies
The standard insulation, #A, seems to suffersignificantly from PD activity and frequency increase
CR material, #B, withstands PD better than #A
PD increase due to PWM voltage waveforms
#B tends to accumulate much more charge than#A (>QM and <s) at frequency up to 50 Hz
#B is worse than #A at 50 Hz
In the presence of PD, the external layer (organicmaterial) is eroded rapidly and the inorganic materialemerges and the enamel change its colour
The inorganic material is easily removed by themechanical stresses (vibrations).
The “Frost Effect”
Let
Type A: only organicmaterials
Type B: organicmaterials filled byinorganic particles
The surface erosion isevident (Type B1) while alocalized BD occurs inorganic enamel (Type A)
Fault Fault
Type A Type B1
Let
Type A: only organicmaterials
Type B: organicmaterials filled byinorganic particles
The surface erosion isevident (Type B1) while alocalized BD occurs inorganic enamel (Type A)
A new class of micro-fillers (Type C) has been developedwhere the inorganic filler chemically combines with theorganic enamel
Type B1
The chemical links delay the mechanical erosionand the insulation life is prolonged
Type C1
The “Frost Effect” combined with electrical stressesdetermine two types of breakdown:
Pinhole Type Massive Type
The two type of BD are related to the time exposition toPD, thus to the local electric field
In this example, BD occurred due to a defect on theconductor where the electric field was enhanced by thecopper protrusions
The “Frost Effect” due to PD is evident looking around thebreakdown site
The three layers are also evident in the picture
Another example of PD erosion
X-Ray spectrometry evidencedthe dominant presence of oxygenin the vicinity of the BD areawhile TiO2 was found around theBD crater
Nano Fillers
To improve the PD resistance of organic enamel, bymeans of the dispersion of nano-metric inorganic fillersare dispersed in the polymer matrix (under investigation)
Polymeric nano-composite: composite material withinorganic fillers having at least one dimension < 100 nm
Polymer NanofillersPolymer Nanofillers
Nanoparticles
Nanotubes, Nanofibres,Whiskers, Nanorods
Nanolayers
Nanocomposite
The filler rate is usually between 1%-10% of weight
The presence of inorganic nano-fillers can alter thedielectric properties of the materials. In particular,
• Permettivity
• Space charge accumulation
• Electrical Tree propagation
• Heat transmission
• etc.
This new technology must be handled carefully to avoidthat improving a property, worsening the others
The nano filler is selected taking into account theproperty to be improved (e.g., the use of nano micaflakes to delay the electric tree growth)
The filler rate is usually between 1%-10% of weight
The presence of inorganic nano-fillers can alter thedielectric properties of the materials. In particular,
• Permettivity
• Space charge accumulation
• Electrical Tree propagation
• Heat transmission
• etc.
This new technology must be handled carefully to avoidthat improving a property, worsening the others
The nano filler is selected taking into account theproperty to be improved (e.g., the use of nano micaflakes to delay the electric tree growth)
EL470+DEL72 (G21)After 22 hours
EL470 (G19)After 3 hours
Tree Growth: bush type
Nano-mica flakes form a wide and complex “labyrinth” wherethe length of the tree-channels are strongly increased andthe breakdown, delayed
EL470+DEL72 (G21)After 22 hours
EL470+MAE (G22)After 20 hours
Anomalous Tree Grouth due tothe barrier-effect of the nanofiller
Nano-fillers are added to improve the resin performances
mainly to withstand the PD erosion
Initialagglomerate
Conventionalcomposit
Intercalatednano-composite
Exfoliatednano-composite
The complete exfoliation ofthe nano-filler generates freecharges inside the insulationworsening e.g., thedissipation factor
RB standard resinN1 1% nano micaN3 3% nano mica
The intercalatedstructure is preferablethe ionic links betweenthe mica flakes arepreserved and no freecharges are introduced
Possible Barrier Effect
Conventional enamel
Inter-turn PDs
Initial stage of aging Further stage of aging
PD induce ablative degradation process leading to thescission of the polymeric chain, the formation of freeradicals and of volatile decomposition productsAdopt nano composite materials (Type C) that show astrong interaction between the nano particles
Nanocomposite enamel
Inter-turn PDs
Initial stage of aging Further stage of aging
PD
CeramicLike layer Increase of
nanofillerconcentrationon the surface
Aggregationforces between
inorganicnanoparticles
Interactions or bondsbetween filler and
carbonaceous residue
Conventional and nano-composite enamel have beenanalyzed and compared
Type A: double layer polyester-imide (PEI)
Type C1: double layer PEI and PEI+Barium SulphateBaSO4 (PEI+nb)
Type C1: double layer PEI and PEI+Silica SiO2 (PEI+ns)
The TBD has been adopted as end-of-life criterium
Tests were performed applying a PWM like wave-form atdifferent voltage levels and temperatures
F.Guastavino et al. “Electrical Aging Tests on Different NanostructuredEnamels Subjected to Severe Voltage Waveforms”, proc.IEEESDEMPED, pp.283-287, Bologna (I), September 2011
The support of Elantas Deatech S.r.l. - Ascoli Piceno – Italy isgreatifully acknowledged
Conventional and nano-composite enamel have beenanalyzed and compared
Type A: double layer polyester-imide (PEI)
Type C1: double layer PEI and PEI+Barium SulphateBaSO4 (PEI+nb)
Type C1: double layer PEI and PEI+Silica SiO2 (PEI+ns)
The TBD has been adopted as end-of-life criterium
Tests were performed applying a PWM like wave-form atdifferent voltage levels and temperatures
F.Guastavino et al. “Electrical Aging Tests on Different NanostructuredEnamels Subjected to Severe Voltage Waveforms”, proc.IEEESDEMPED, pp.283-287, Bologna (I), September 2011
The support of Elantas Deatech S.r.l. - Ascoli Piceno – Italy isgreatifully acknowledged
Test Set-Upoven
VAWADIT
ArbitraryWaveformGenerator
Twisted pair
oven
VAWADIT
ArbitraryWaveformGenerator
Twisted pair
PWM+peaksvoltage
waveform
Linear amplifier: 10 Hz– 3 MHz bandwidth atthe considered voltage
level, 60 dB gain
Temperaturetest: 150°C;120°C; 90°C;
60°C
The average time to breakdown (Tbd) is collected andrelated to the test voltage amplitude via the inversepower law:
Tbd = A (Vpp)-n
1000
10000
1000 10000 100000 1000000
Vte
st[V
]
Tbd [s]
PEI
PEI+ns
PEI+nb
1000
10000
1000 10000 100000 1000000
Vte
st[V
]
Tbd [s]
PEI
PEI+ns
PEI+nb
The fomation of nanostructured ceramic-like layer has beenobserved for many ablative processes:• Burning (Giannelis et al., Gilman et al) .• Thermo-oxidative degradation (Mulhaupt. et al., Zanetti et al., Camino et al.)
• Exposure to combustion gases (Vaia et al.)
The described processes have been massively evidencedfor Polymer layered silicate nanocomposites, but similarbehavior has been observed also in the case of polymer-SiO2 nanocomposites (Wu et al. 2005, Wang et al. 2006, also according to the work ofVaia).
Ceramic char formation during ablation
The described processes have been massively evidencedfor Polymer layered silicate nanocomposites, but similarbehavior has been observed also in the case of polymer-SiO2 nanocomposites (Wu et al. 2005, Wang et al. 2006, also according to the work ofVaia).
Qualitative Surface Analysis
Electricalaging tests
PDsactivity
Enamelerosion
Qualitativesurface analysis
Simple opticalmicroscope
Degradation area dimensionsfor conventional enamel after
electrical aging at 150°C
Degradation area dimensions fornanocomposite enamel after
electrical aging at 150°C
Degradation area dimensions are wider in the case ofconventional enamel than in the case of nanocomposite one
Degradation area dimensionsfor conventional enamel after
electrical aging at 60°C
Degradation area dimensions fornanocomposite enamel after
electrical aging at 60°C
Diminishing the temperature level, the eroded areadimensions is less wide
PEI+nb twisted pair
Before aging test
After aging test at 4.6 kV
PEI+ns twisted pairs
Before aging test After aging test at 4.6 kV
Comparison Between PEI+nb and PEI+ns
After aging test at 4.6 kVAfter aging test at 4.6 kV
PEI + nb PEI + ns
Thermal Ageing
Applying the Arrheniusmodel to the obtained
life timesLinearizing Life Curves
60 90 120 150
T [°C]
1000
10000
100000
1000000
0,00220,00240,00260,00280,00300,0032
1/T [1/K]
Tbd
[s]
Conventional
Nanocomposite
60 90 120 150
T [°C]
1000
10000
100000
1000000
0,00220,00240,00260,00280,00300,0032
1/T [1/K]
Tbd
[s]
Conventional
Nanocomposite
60 90 120 150
T [°C]
1000
10000
100000
1000000
0,00220,00240,00260,00280,00300,0032
1/T [1/K]
Tbd
[s]
Conventional
Nanocomposite
60 90 120 150
T [°C]
1000
10000
100000
1000000
0,00220,00240,00260,00280,00300,0032
1/T [1/K]
Tbd
[s]
Conventional
Nanocomposite
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Tbd
[s]
0
2000
4000
6000
8000
1000030000
45000
60000
75000
90000
105000
120000
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Tbd
[s]
0
2000
4000
6000
8000
1000030000
45000
60000
75000
90000
105000
120000
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Tbd
[s]
0
2000
4000
6000
8000
1000030000
45000
60000
75000
90000
105000
120000
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Conv.150°C
Nano150°C
Conv.120°C
Nano120°C
Conv.90°C
Nano90°C
Conv.60°C
Nano60°C
Tbd
[s]
0
2000
4000
6000
8000
1000030000
45000
60000
75000
90000
105000
120000
Data scatter is generally low; the minimum life time value obtained testing thenano-composite enamel at 150°C is considerablylonger than the maximum time obtained testing theconventional enamel at 60°C
Micro+Nano Composites
A combined use of micro and nano fillers has been alsoinvestigate to
improve different properties of the composite materials(thermal and mechanical in addition to PD resistance)
guarantee novel properties
Schematicrepresentation of PDerosion process due toPD for micro andmicro+nano compositematerials
Sample of micro-silica(60%wt) and nano-silica (5%wt) in epoxymatrixMicro-silica: black areaNano-silica: small withedots
CIGRE Working Group, “Characterization of Epoxy Microcomposite andnanocomposite Materials for Power Engineering Applications”, IEEEEl.Ins.Magazine, Vol.28, pp.38-51, March 2012
It is possible to enhance the resistance to the action ofsurface PDs of organic insulating enamels used formagnet wire insulation by nano-structuration.
The application of nano-composite enamels is not aPanacea for inverter driven motor insulation:
1.Many matrix-filler combination may not lead to thedesired results; careful study of the chemical-physicalinteractions and degradation mechanisms
2.Nano-structuration does not prevent the inception ofPDs; rather it slows down the degradation of theenamel
It is possible to enhance the resistance to the action ofsurface PDs of organic insulating enamels used formagnet wire insulation by nano-structuration.
The application of nano-composite enamels is not aPanacea for inverter driven motor insulation:
1.Many matrix-filler combination may not lead to thedesired results; careful study of the chemical-physicalinteractions and degradation mechanisms
2.Nano-structuration does not prevent the inception ofPDs; rather it slows down the degradation of theenamel
Required further research investigation:1.Chemistry and physics of the degradation of nano-
composite enamels subjected to PDs2.Polymer – inorganic nano-particles interactions3.Interactions between nano-composite enamels and
secondary insulation (conventional or nano-structuredimpregnation resins)
4. Micro-nano composites
Required further research investigation:1.Chemistry and physics of the degradation of nano-
composite enamels subjected to PDs2.Polymer – inorganic nano-particles interactions3.Interactions between nano-composite enamels and
secondary insulation (conventional or nano-structuredimpregnation resins)
4. Micro-nano composites
The concepts of dielectric strength, Weibull distribution offailure times, lifetime and voltage endurance coefficientare the basis for the design of highly reliable insulationsystems in electrical apparatus
ASD introduced a new type of electrical stress arisingfrom high frequency harmonics due to repetitive voltageimpulses and motor-cable-converter impedancemismatch
Over-voltages and uneven voltage distribution along thewinding causes overstress mainly in inter-turn insulation
Over-voltages can cause PD that become the dominantageing factor mainly in Type I materials
Corona resistant materials (micro, nano composites) havebeen developed
Modeling for Insulation Design
The concepts of dielectric strength, Weibull distribution offailure times, lifetime and voltage endurance coefficientare the basis for the design of highly reliable insulationsystems in electrical apparatus
ASD introduced a new type of electrical stress arisingfrom high frequency harmonics due to repetitive voltageimpulses and motor-cable-converter impedancemismatch
Over-voltages and uneven voltage distribution along thewinding causes overstress mainly in inter-turn insulation
Over-voltages can cause PD that become the dominantageing factor mainly in Type I materials
Corona resistant materials (micro, nano composites) havebeen developed
It is necessary to study and model the lifetime behavior of new CRmaterials to design properly the insulation taking into account theASD specific stresses
The dominant aging factors
Some quantities extracted from the distorted voltagewaveforms and correlated with aging are introduced. Let:Some quantities extracted from the distorted voltagewaveforms and correlated with aging are introduced. Let:
N
hnfh thVtv
1
)sin()(
The Fourier series of the non-sinusoidal voltage supply.
N
hnfhf thVh
dt
tdv
1
)cos()(
The rms of the voltage variation is defined as:
If we consider the rms value of a 50 Hz sinusoidal voltagehaving the same amplitude of the fundamental (V0=V1),then:
N
hh
f
rms
Vhdt
tdv
1
22
2
)(
10
,50
0
2
)(V
dt
tdv
rmsHz
Considering their ratio where
10
,50
0
2
)(V
dt
tdv
rmsHz
N
hh
fs hK
1
22
0
1V
Vhh
Ks is the rms value of the derivative of thedistorted waveform and it is related to its RT
*1rms
rmsrms V
VK
*1P
PP V
VK where
VP and Vrms are the peak and rms valuesof the distorted waveform, V1
* is the reference voltage (V1P*=√2 ⋅
V1rms),
Additional parameters, related to over-voltages, can bedefined as:
*1rms
rmsrms V
VK
where VP and Vrms are the peak and rms valuesof the distorted waveform, V1
* is the reference voltage (V1P*=√2 ⋅
V1rms),
The Joule, Wj, and the dielectric, Wd, losses for a windinghaving a phase-to-ground capacitance C, can be writtenas:
N
h
haaJ V
VhVCrkW
1
2
1
221
221 )(
N
h
hd V
VhCVW
1
2
1
211 )(tan
Where ra is the resistance of the equivalent capacitor ka is a constant tan is the loss factorThe temperature rise is then given by:
Where ra is the resistance of the equivalent capacitor ka is a constant tan is the loss factorThe temperature rise is then given by:
thdJ RWW )( where Rth is the thermal resistance of the capacitorUsing the PWM technique, temperature increasesof about 10 to 20 K°
The Dominant Ageing Factors
With PD: It has been shown that PD is the dominantageing factor particularly at high pulse rate and frequency
Thus, the insulation system must be designed to workbelow the PDIV
PDIV depends on the adopted insulation
Without PD: Below the PDIV, the ageing mechanisms arevery different (related to RT, RR, temperature….)
Neglecting the interactions between factors as a firstapproximation, the simplest equation that can be used,based on an inverse power model, is:
The Dominant Ageing Factors
With PD: It has been shown that PD is the dominantageing factor particularly at high pulse rate and frequency
Thus, the insulation system must be designed to workbelow the PDIV
PDIV depends on the adopted insulation
Without PD: Below the PDIV, the ageing mechanisms arevery different (related to RT, RR, temperature….)
Neglecting the interactions between factors as a firstapproximation, the simplest equation that can be used,based on an inverse power model, is:
srmsP ns
nrms
nP KKKLL 0
where L is insulation lifetime, and L0, np, nrms, nsare adjustable parameters
KP, Krms, and Ks are further analyzed statistically, e.g., byusing the Standardized Pareto Chart (SPC) and the MainEffect Plot (MEP)
ssrmsrmsPp KnKnKnL
Lloglogloglog
0
Peak voltage is clearly the most influential factor of lifetime,followed by rms and voltage slope
nba NBALL 0
The experimental data suggested that an inverse powermodel, in the form of
can be applied to correlate life- time and aging factors, inparticular P2P voltage and temperature, that is, in logform
loglogloglog bVaLL PPD
where L=lifetime; VPP=P2P voltage; and a, b and L0 areparameters calculated through multivariable linearregression
Again, the jumpvoltage is still themost influential factorof lifetime
where L=lifetime; VPP=P2P voltage; and a, b and L0 areparameters calculated through multivariable linearregression
Peak-to-peak voltage has been recognized as dominantageing factorThe inverse power model correlating the lifetimes atdifferent stress values is slightly modified, that is
nppVLL 0
Life Modeling: a simplified version
The pulse RR of the applied voltage is also important, withlifetime decreasing with increasing RR. If Lf and L1 are thelifetimes at f (= 10 kHz) and f1 (= 50 Hz), respectively then
The pulse RR of the applied voltage is also important, withlifetime decreasing with increasing RR. If Lf and L1 are thelifetimes at f (= 10 kHz) and f1 (= 50 Hz), respectively then
)( 11 f
fLLf
where the exponent γ is estimated experimentally
These assumptions allow the lifetime of an insulationsystem under impulse conditions to be estimated usingdata obtained under sinusoidal voltage testing. Moreover:
The most important stressing factors are fundamentalfrequency, the RR and peak-to-peak voltage amplitude
The overvoltage is adiabatic
The system is operated in the stress range within which thepredominant degradation mechanism does not change duringageing
VEC is frequency independent, i.e., no significant frequencydependence of the number of impulses or voltage cycles beforefailure is observed. This corresponds to = 1 which isapproximately true for composite organic/inorganic insulation
These assumptions allow the lifetime of an insulationsystem under impulse conditions to be estimated usingdata obtained under sinusoidal voltage testing. Moreover:
The most important stressing factors are fundamentalfrequency, the RR and peak-to-peak voltage amplitude
The overvoltage is adiabatic
The system is operated in the stress range within which thepredominant degradation mechanism does not change duringageing
VEC is frequency independent, i.e., no significant frequencydependence of the number of impulses or voltage cycles beforefailure is observed. This corresponds to = 1 which isapproximately true for composite organic/inorganic insulation
If these conditions are satisfied and the measured lifetimeat test frequency f1 is L1, then the estimated lifetime L2 attest frequency f2 is given by:
2
112 f
fLL
It follows that
if lifetime line 1corresponds to testfrequency f1,
lifetime line 2 for testfrequency f2 (f2 = 10 f1) isobtained by translatinglifetime line 1 one decadehorizontally (arrow A)
It follows that
if lifetime line 1corresponds to testfrequency f1,
lifetime line 2 for testfrequency f2 (f2 = 10 f1) isobtained by translatinglifetime line 1 one decadehorizontally (arrow A)
to maintain the f1 lifetime, the applied stress at f2should be reduced as shown by arrow B
Combining the two simplified models
2
1
2
1,, )(
1122 f
f
U
ULL n
ufuf
where Lf1,u1 is the lifetime at frequency f1 and voltage U1
Lf2,u2 is the lifetime at frequency f2 and voltage U2
n is the VECdata can be generated for any desired frequency, e.g., thefundamental for motor drives, based on measured lifetimesfor appropriate insulation systems at f1 = 50 or 60 Hz
Experimental evidence validates this simplified approach forimpulse voltages up to 1 kHz because the variation of theVEC n with frequency is negligible
where Lf1,u1 is the lifetime at frequency f1 and voltage U1
Lf2,u2 is the lifetime at frequency f2 and voltage U2
n is the VECdata can be generated for any desired frequency, e.g., thefundamental for motor drives, based on measured lifetimesfor appropriate insulation systems at f1 = 50 or 60 Hz
Experimental evidence validates this simplified approach forimpulse voltages up to 1 kHz because the variation of theVEC n with frequency is negligible
At higher frequencies a decrease of n is observed, even forinorganic/organic insulation
The dependence of n with frequency can be modeled, butonly by introducing further parameters in the model.
Design Criteria
PD is a dominant deterioration phenomena that leads topremature BD of the insulation. Using conventional enamel,the electric stress must be below the PDIV
a moderate PD activity can be accepted when CRcomposite organic/inorganic insulating material is adopted
In the absence of PD, the peak of the distorted voltagewaveform and its repetition rate are the most importantageing factors
Modeling the long-term behavior is feasible in the firstapproximation
Detailed evaluation of the new materials through long-term voltage endurance tests is still strongly recommended,to maximize the reliability of the insulation system
PD is a dominant deterioration phenomena that leads topremature BD of the insulation. Using conventional enamel,the electric stress must be below the PDIV
a moderate PD activity can be accepted when CRcomposite organic/inorganic insulating material is adopted
In the absence of PD, the peak of the distorted voltagewaveform and its repetition rate are the most importantageing factors
Modeling the long-term behavior is feasible in the firstapproximation
Detailed evaluation of the new materials through long-term voltage endurance tests is still strongly recommended,to maximize the reliability of the insulation system
A. Cavallini, D. Fabiani, G.C. Montanari, “Power Electronicsand Electrical Insulation System – Part 1: PhenomenologyOverview”, IEEE Electrical Insulation Magazine, Vol. 26, pp. 7-15, May-June 2010
A. Cavallini, D. Fabiani, G.C. Montanari, “Power Electronics andElectrical Insulation System – Part 2: Life Modeling for InsulationDesign”, IEEE Electrical Insulation Magazine, Vol. 26, pp. 33-39, July-August 2010
A. Cavallini, D. Fabiani, G.C. Montanari, “Power Electronics andElectrical Insulation System – Part 3: Diagnostic Properties”, IEEEElectrical Insulation Magazine, Vol. 26, pp. 30-40,September-October2010
References:
A. Cavallini, D. Fabiani, G.C. Montanari, “Power Electronicsand Electrical Insulation System – Part 1: PhenomenologyOverview”, IEEE Electrical Insulation Magazine, Vol. 26, pp. 7-15, May-June 2010
A. Cavallini, D. Fabiani, G.C. Montanari, “Power Electronics andElectrical Insulation System – Part 2: Life Modeling for InsulationDesign”, IEEE Electrical Insulation Magazine, Vol. 26, pp. 33-39, July-August 2010
A. Cavallini, D. Fabiani, G.C. Montanari, “Power Electronics andElectrical Insulation System – Part 3: Diagnostic Properties”, IEEEElectrical Insulation Magazine, Vol. 26, pp. 30-40,September-October2010
In form wound MV and HV rotating machines, theground-wall and strand insulation, based on mica-tapesand mica-films, respectively, withstand to the PD activity
Turn insulation is stressed by the uneven voltagedistribution and can be designed according to the abovementioned composite materials and criteria
Design of the Stress Grading for HV Machines
The end-arm stress grading isthe weak point when a machinedesigned for 50/60 Hz is fed bypulsating of PWM supply. Thispromoted investigations to:
select proper materials
design properly the stressgrading
The end-arm stress grading isthe weak point when a machinedesigned for 50/60 Hz is fed bypulsating of PWM supply. Thispromoted investigations to:
select proper materials
design properly the stressgrading
Problemi nei sistemi digradatura
Solo nei motori form-wound (MT)• Il campo elettrico tangenziale
aumenta all’aumentare delcontenuto in frequenza dellatensione (f> 2 kHz si hanno scarichein testata)
• Il campo elettrico cala all’aumentaredella conducibilità della gradatura
Solo nei motori form-wound (MT)• Il campo elettrico tangenziale
aumenta all’aumentare delcontenuto in frequenza dellatensione (f> 2 kHz si hanno scarichein testata)
• Il campo elettrico cala all’aumentaredella conducibilità della gradatura
Media conducibilità Alta conducibilità
Campo elettricodi scarica Campo elettrico
di scarica
Typically:
• Insulation with no stress grading system(normally for Vn ≤ 4kV)
• Insulation systems with anticorona coating within theslot ( 4kV ≤Vn ≤ 6kV)
• Insulation with stress grading system (Vn ≥6kV)
Typically:
• Insulation with no stress grading system(normally for Vn ≤ 4kV)
• Insulation systems with anticorona coating within theslot ( 4kV ≤Vn ≤ 6kV)
• Insulation with stress grading system (Vn ≥6kV)
Due to material discontinuity, high values of electricgradient affect the surface of the coil at the edge of the slotgrading tape thus generating tangential surface discharges
The stress grading is designed solving the field equation
0)*(
U
tUcc
and its solution allows to draw the electric field outsidethe magnetic core
)cosh(
)(cosh1)(
kL
LxkVxU a
)cosh(
)(cosh1)(
kL
LxkVxU a
)sinh(
)(sinh)(
kL
LxkkVxE a
Hot spots due to PD, can be discovered considering theair breakdown strength (e.g., 2.3 kV/mm) and theelectric field gradient
Stress grading materials are characterized by theirresistance that can be constant or electric field dependent:
)exp( 3/20 nEs
High values of n and low values of 0 increase the gradingeffect
The stress grading is designed considering the number oflayers and their coating length outside a slot portion
After the material selection (0, n), the space distributionof the electric field is determined using FEM softwaretools considering the machine geometry and the ground-wall insulation characteristicsThe correct choice of the number of layers and theirlength is evaluated checking the electric gradient (belowof PDIV) and hot spots
High values of n and low values of 0 increase the gradingeffect
The stress grading is designed considering the number oflayers and their coating length outside a slot portion
After the material selection (0, n), the space distributionof the electric field is determined using FEM softwaretools considering the machine geometry and the ground-wall insulation characteristicsThe correct choice of the number of layers and theirlength is evaluated checking the electric gradient (belowof PDIV) and hot spots
Stress grading is currently designed for 50/60HzapplicationsAssuming the 0 reference the edge of the slot grading, thepotential and the electric field distribution can be derivedand analyzedThe different behavior of a single layer stress grading withconstant and exponential resistivity, is show
But the potential distribution is strongly related tofrequency of the applied stress.Stress grading designed for ac is not able to operate athigher frequencies
Distanza x dal termine del ricoprimento conduttivo [mm]
SIM.1.2.1.A SIM.1.2.1.B
Mat.D
Mat.A
Vmax [V]
16 kV # 1,25 MHzMod.1
s = 2,7 mm
1
4
2
3
Experimental Validation
Two different stress grading configurations designed forstandard (50 Hz) and PWM supply voltages have been testedby means of PD measurements
Both frames were supplied by HV rectangular wave-shape
PD measurements were performed using an antenna probeable to record PD pulses and the fundamental wave-shapeadopted as the phase reference of PD
Moving the antenna probe, PD were localized at the edge ofthe 50Hz stress grading while only signals due tocommutations were recorded on the other frame
Thus confirming the validity of the stress grading design
Besides the advantages in using ASD, new problemsrose due to the significant harmonic content of the powersupply and the over-voltages generated by mismatchimpedances in inverter/cable/drive connection
The insulation is subjected to increased electric,thermal and mechanical stresses and its life is shortened
Additional stresses, typical of ASD, have beenexamined
New composite materials (micro, nano fillers) havebeen proposed
Specific test methods are developed
Specific standards are under discussion
Conclusions
Besides the advantages in using ASD, new problemsrose due to the significant harmonic content of the powersupply and the over-voltages generated by mismatchimpedances in inverter/cable/drive connection
The insulation is subjected to increased electric,thermal and mechanical stresses and its life is shortened
Additional stresses, typical of ASD, have beenexamined
New composite materials (micro, nano fillers) havebeen proposed