Flashover voltage of silicone insulating surface covered by water droplets under AC voltage

(1)

Contents lists available atScienceDirect

Electric

Power

Systems

Research

j o u r n a l h o m e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / e p s r

Flashover

voltage

of

silicone

insulating

surface

covered

by

water

droplets

under

AC

voltage

F. Aouabed

a

_,

_A.

_Bayadi

a,∗

_,

_R.

_Boudissa

b

a_Département_{Electrotechnique,}_Université_Ferhat_Abbas_de_Sétif_1,_Sétif,_Algeria

b_Labo_Génie_Electrique,_Université_A._Mira,_Béjaia,_Algeria

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received26May2015

Receivedinrevisedform

13September2016 Accepted10October2016 Keywords: Contamination Flashover Testing

Siliconerubberinsulators

Surfacewettability

Waterdroplets

a

b

s

t

r

a

c

t

Dischargescausedbywaterdropletsonthesurfaceofpolymericinsulatorscanaffectthelong-term reliabilityofthecomponentbyloweringthesurfacehydrophobicityboostingsurfacedischarges.The mainobjectiveofthisworkistoquantifytheeffectofdifferenttypesofwaterdropsarrangements,their positionanddrybandswidthontheﬂashovervoltageofthesiliconeinsulatingsurfacewithnon-uniform electricﬁeldsystems.ThetestsweredoneonarectangularsampleunderACvoltage.Waterdropletswith differentconductivitiesandvolumeswereplacedonthesiliconerubbersurfacewithamicropipette.A rod-rodelectrodesystemisused.

Thefindingsofthisworkindicatethattheperformanceofthesamplesdecreaseswiththepresenceof waterdropsontheirsurfaces.Further,theseexperimentalfindingsshowthatthereisalimitingnumber ofrowsfromwhichtheflashovervoltageoftheinsulationisminimalandconstant.Thisminimumisa functionofthedistancebetweentwosuccessiverows.

Finally,itisconcludedthatthesystemwithstandvoltageincreaseswhentherowofdropletsonthe electrodeaxisisremoved.

1. Introduction

Nowadays,siliconerubberinsulationmaterialsarewidelyused inhighvoltageoutdoorinsulationsystemsastheycancombat pol-lutionﬂashoverproblems.

Thedifferenceinpollutionflashoverperformance ofsilicone rubberandotherinsulatingmaterialsisduetothewaythatwater wetstheirsurfaces.Itresidesasdiscretedropsonsiliconerubber, andthemechanismofflashoverisduetothebreakdownofthe airbetweenthewaterdropsandthedeformationofthesedrops inthedirectionoftheelectricfieldwhichbringstheinsulationto degradationandfailure.

Duringthepast25yearspolymericmaterialshaveemergedasa viableoptiontoporcelainand/orglassforoutdoorinsulation[1,2]. Polymericinsulatorsareincreasinglybeingusedinboththe dis-tributionandtransmissionsystemsbecauseoftheirverystrong resistancetothecontamination,theirlightness,theirmechanical resistanceand theirverygood wettability.The deterioration of

∗ Corresponding author at: Département d’Electrotechnique, Faculté de

Technologie, Université Ferhat Abbas Sétif 1, Algeria. Tel.:+213665303202;

fax:+21336611211.

E-mailaddresses:fatihaess@yahoo.fr(F.Aouabed),abayadi@yahoo.fr,

abayadi@univ-setif.dz(A.Bayadi),raboudissa@yahoo.fr(R.Boudissa).

insulatorsurfaceisoneofmainproblemstothesafetyand reli-abilityofelectricsystems.Thisisduetoenvironmentalconditions (light rain,morning dew), which significantly affect its perfor-mance.Itisknownthatwaterdropletsmaycause,underapplied electricfield,deteriorationofthesurfaceofanon-ceramic insula-toreveninconditionsoflowpollution.Thisisduetothefactthat waterdropletsonapolymersurfacelocallyincreasetheapplied electricfield. Localfield intensificationswilllead topartial dis-chargesand/orlocalizedarcs.Thesepartialdischargesdestroythe hydrophobicityandcausethedegradationoftheinsulatorswhich canplayanimportantroleinlong-termperformance[9–13].The influenceofvariousparametersonthebehaviorofwaterdroplets onpolymericsurfacesunderhighelectricfieldshasbeenthesubject ofseveralinvestigations[3–13].Themainobjectivewastoincrease thealternatingelectricfield.Karady[5],forexample,presentedthe resultsofanexperimentalinvestigationwhichprovidesabetter understandingofthephenomenaleadingtoflashover.Itwasshown thatthehydrophobicnatureofsiliconerubbersurfaceresultsina flashovermechanismdifferentfromthatofporcelaininsulators. Changesinsurfaceresistanceinducedbydischargeactivityonwet andcontaminatedsurfacehavebeenidentifiedasoneofthesources offlashoverofsiliconeinsulators.Phillipsetal.[10]publishedthe resultsofaresearchthatcorrelatestheinsulatorsagingwithcorona activitiesfromwaterdroplets.Usingsmall-scaleexperiment,itwas shownthatWaterdropsonthesheathsurfacesofSiRinsulatorcan http://dx.doi.org/10.1016/j.epsr.2016.10.025

(2)

producecorona,andtheﬁeldnecessarytoproducecoronadepends onthedropletssizeandthesurfacehydrophobicity.

Swift[14] studiedtheoreticalandexperimentalinvestigation ofdropletsonthesurfaceinsulatorfromtheviewpointofwater triggeredelectricalbreakdownofanair-dielectricinterface.Itwas shown that greatly reducing thehydrophobicity of the surface by sparking the ﬂashover voltage, but recovery to the fully-hydrophobicvaluetakesonlyafewminutes.andthatforthefully hydrophobiccase,ﬂashoveristriggeredbygrossdistortionofthe waterdroplet.

InRef.[15],theauthorshavepresentedanexperimentalstudy ontheproblemsarisingfromtheapplicationofuniformacelectric ﬁeldsonwaterdropletsdepositedonpolymersurfaces.

Differentpolymericmaterialswereused.Itwasshownthat vari-ousparameterssuchaswaterconductivity,dropletvolume,droplet positioningand polymer surface roughness affectthe ﬂashover voltage.Itwasalsoreportedthatthepositioningofthedroplets playsagreaterroleindeterminingtheﬂashovervoltagethanthe dropletvolume.

InRef.[16],testshavebeendoneonwaterdropletsof differ-entconductivitiesandvolumesdepositedonthesurfaceofsilicone rubber.Factorsaffectingthecoronadischargeofwater droplets wereanalyzedbycalculatingtheelectricﬁeld.Theauthorsreported thatwaterdropletsincreasetheelectricﬁeldandcancausecorona discharge.Theyalsoobservedthatthedropletvibratesandliedown tothepositiveelectrodeinawavesynchronismwiththeapplied voltagefrequency.

Fernandoetal.[17]studiedthebehaviorofleakagecurrenton compositeinsulatorsofdifferentmaterials.Theyconcludedthat theleakagecurrentiscapacitiveinnatureandoftenhasa sinu-soidalshape.Whenthehydrophobicityislost,theleakagecurrent becomesmoreresistive,withpeaksduetodischargesindrybands. Lopesetal.[20] measuredpartialdischarge(PD)fromwater dropletsonasiliconerubberinsulatingsurfaceinanacfield.They haveshownthatthepresenceofwaterdropletsonasiliconerubber surfaceproducesanelectricfieldenhancement.Thefield enhance-mentfactordependsonthesizeandnumberofdroplets.Theyalso observedthattheelectrostaticforces changethedropletshapes andspreadtheminthefielddirection.

Souzaetal.[21]investigatedthecoronainceptionandits rela-tion withpolymer surface conditions. They concludedthat the associationwaterdropletsandpollution,enhancetheelectricﬁeld andcanleadtocoronadischargesandeventuallytofailures,under heavycontaminatedregions.

Phillips et al. [22], for example, presented summarized the ﬁndingsoftheextensiveresearchcarriedoutbythesetwo orga-nizations (EPRIand STRI)todeterminea practical limit forthe permissiblee-ﬁeldoninsulatorsurfacesfordesignpurposes.

However,totheauthors’bestknowledge,Afewtheoreticaland experimentalworksstudyingtheinﬂuenceofthecombinationthe dropletsarrangementswithdrybandsonpolymericsurfacehave beendone[18].

Toformabetterviewofthebehaviorofwaterdropletsunder anon-uniformelectricfield,wecarriedoutseveralexperimental testsinthehighvoltagelaboratory.Weinvestigatedtheinfluence ofthenumberofwaterdropletsrowsandthedistancebetween them.Thispapersummarizesfindingsofexperimentswhichallow quantifyingtheseeffectsontheflashovervoltage.

2. Experimentalsetupandtestprocedure

Thewithstandvoltagemeasurementandvisualizationofthe airgapdisruptionphenomenonwereconductedusingequipments depictedinFig.1.ThetestcircuitconsistsofaHVtransformer(Tr) havingamaximumsecondaryvoltageof140kV.Acontrolunit(CU)

Fig.1. Laboratorytestsetup.

(CU:transformercontrolunit;Tr:HVtransformer;Cam:camescope;PC:personal

computer;To:testobject;C1,C2:capacitivedivider;Ra:resistance;V:digitalpeak

voltmeter).

forautomaticormanualspeedrampcontrol.Adigitalpeak volt-meter(V)atthelow-voltagearmofacapacitivedivider(C1,C2)

wasusedforvoltagemeasurements.Acurrent-limitingresistance (Ra)wasconnectedinserieswithtestobject(To).Thedevelopment

oftheelectricaldischargealongthesamplesurfacewasvisualized andrecordedfrominceptiontofullﬂashoverusingavideocamera system

Theairgapsystemhastwopointelectrodes.Thehigh-voltage andearthelectrodesconsistofacylindricalstainlesssteelrod5mm indiameter.Theyareterminatedbyaconicaltiphavinganangle of60◦andradiusofcurvatureof0.5mm.

TherodsareﬁxedontheholesoftwoPVCtubes.

The test specimen is a plate-shaped silicone rubber having 120mm in length,80mm in width and 5mm in thikness.The pollution solutioncomprises saltand distilledwater. Thewater droplets havebeendepositedonthehydrophobicsurfaceusing amicropipette.Thevolumeofthesedropletscanbeobtainedby tuningthemicropipetteatthedesiredsize.Inthiscase,thevolume ofwaterdropletsdepositedis40␮l.Thevolumeconductivityv

ofthepollutingsolutionwasdirectlymeasuredbyamobileprobe volumeconductimeterandwasfoundtobeabout180␮S/cm.

Theairgapdistancewastakenequalto6cm.Twentyﬁvetests werecarriedoutandvoltagestepsVofapproximately5%ofUc

wereused.Thesetestparametersarewellwithintherecommended valuesforsuchstatisticalmeasurementswherethenumberoftests n shouldbebetween20and 60 testsand thevoltagestep V between1%and 10% oftheﬂashovervoltage. Beforeeach new test,thesampleiscleanedandrinsedwithwaterthendriedusing papertissue.Thencleanedwiththeisopropanolalcohol.Foreach calculatedmeanvalue,byapplyingcorrectionfactors,adisruptive dischargevoltagemeasuredingiventestconditions(temperature T, pressure P,humidityH) isconverted totheequivalentvalue underthestandardreferenceatmosphericconditions(T0=20◦C,

P0=101.3kPa,H0=11g/m3)[19].Thegroupofwaterdropletsis

characterizedbyacoupleofvariable(ar,dg).Thedistancearisthe widthbetweentwosuccessiverowsanddgisthedistancebetween twosuccessivecolumnsofwaterdropletsasshowninFig.2. 3. Resultsanddiscussion

Toinvestigatetheinﬂuenceofwaterdropletontheﬂashover voltageofasiliconerubberinsulatingsurfaceaseriesof experi-mentshavebeenconductedasfollows:

3.1. Effectofwaterconductivity

Fig.3showstheinﬂuenceofthewaterdropletsconductivityon themeanvalueoftheinsulatorﬂashovervoltagefortwodifferent numbersofrows.

(3)

Fig.2. Distributionofwaterdropletsonthesample.

Fig.3.Inﬂuenceofthewaterdropletsconductivityontheinsulatorﬂashover

volt-age.

Theobtainedresultsshowthatthechosenvaluesofrowsdoes notsigniﬁcantlyaffectthecharacteristicUc=f(v).

Furthermore, as can be seen, the average ﬂashover voltage decreasesbyapproximately38%withthepollutionconductivity untilalimitvalueofabout2.5mS/cmbeyondwhichthedielectric

strengthofthesystemremainsconstant.Thisismainlyduetothe factthatfromthislimit,thewaterbecomesconductive.

3.2. Numberofrowsofwaterdroplets

Inthispartofourinvestigations,experimentswereperformed withoutanydropletsbetweentheelectrodes.Thiswasdonein ordertohavereferencevaluesoftheflashovervoltage.Inorder tounderstandtheinfluenceofdropletsnumberofrowsbetween theelectrodes,thelatterisvariedfrom1to11.Thevolume con-ductivityismaintainedconstantatabout180␮S/cm.Theobtained resultsareshowninFig.4.Theobservationswiththevideocamera systemhaveshownthattheflashoverprocessfromtheinception tofullflashoverisdescribedessentiallyasfollows:

a)First,thegroupofwaterdropletsdepositedonthesample sur-face,beforeitsenergization,isshowninFig.4a.Themiddlerow coincideswiththeelectrodesaxisanditsextremedropletsgot incontactwiththeirpoints.

b)In the second phase, Whenan electric field is applied, It is observedthatthewaterdropletsshapechangealongthe elec-trode axis leading to the decrease of the ignition distance (Fig.4b)andthesharpedgeofthewaterdropletsatthetripleline togetherwiththeoppositeelectrodeformanon-homogeneous fieldconfigurationwhichisthebasiccausetohavethestreamer inceptiononthesamplesurface.

c)Next,forthereasonthattheelectric-ﬁeldstrengthneartheHV electrodeisstrongenough,electricaldischargeisestablished alongtheelectrodeaxisasshowninFig.4c,

d)Finally,thetestedsampleafterthefullﬂashoverisshownin Fig.4d.

Thevariationofﬂashovervoltageasafunctionofwaterdroplets numberis illustratedinFig.5.It canbeseenthatthepresence of water droplets covering the hydrophobic insulating surface energizedwithalternatingvoltagecausesthereductionofits per-formance.Indeed,theﬂashovervoltagedecreasesasthenumberof rowsincreases.Thiscanbeexplainedbythedecreaseofthe igni-tiondistancedue,inonehand,tothespaceoccupiedbyvolume ofwaterdropletsbeforedeformationandintheotherhand,tothe theirstretchingalongthepathfollowedbytheelectricdischarge.

Takingintoaccounttheseresultsitcanbeseenthattheﬂashover voltagedecreasesrapidlytoalimitvalueofthenumberofrows

(4)

Fig.5.Flashovervoltagevsthenumberofrowsdropletsfordifferentdistances

betweenthem.

Fig.6. Relativereductionofﬂashovervoltageasafunctionofrowsnumber.

fromwhich,itbecomesconstant,themaximumvariationofelectric performanceisestimatedto35%.Thevalueofthislimitisa func-tionofthedistancebetweentwosuccessiverows.Furthermore, whenthedistancearisdecreasedalongthecreepagedistanceof thesample,asigniﬁcantdecreaseinitsperformanceisobservedas well.

Fig.6 showsthe relativereduction of ﬂashovervoltageas a functionofrowsnumber.Wecanseethattherelativereduction increasesrapidlytoalimitvalue(5rows)ofthenumberofrows fromwhich,itbecomesrelativelyconstant,whichisinaccordance withtheresultspreviouslyobtained.

The relative reduction betweenﬂashover voltage and water dropletsrowscanbeobtainedasfollows:

U =100%×(U0−Ui)/U0 (1)

whereU0,Uiaretheﬂashovervoltagecorrespondingtothecase

withoutandwithwaterdropletsrowsrespectively.iisvariedfrom 1to11.

Quantitatively,Table1summarizestheresultsobtainedforthe maximumrelativevariationofUCwheniistakenequalto0and

5respectively.Themaximumﬂashovervoltageoftheairgapwith ﬁverowsisabout35%lowerthanthatobtainedinthecaseofadry cleanatmosphere.Itshouldbenotedthatforthelastvalueinthe table,only3rowswereusedduetothedimensionsofoursamples.

Table1

Relativereductionofﬂashovervoltagewithnumberofrows.

Numberofrows Uc(kV) U(%) ar=1cm,dg=1cm 0 41.7 35% 5 27.31 ar=1cm,dg=2cm 0 41.7 14% 5 35.94 ar=2cm,dg=2cm 0 41.7 8% 5 38.43 ar=3cm,dg=2cm 0 41.7 5.35% 5 39.47 ar=4cm,dg=2cm 0 41.7 5.5% 3 39.4 0 1 2 3 4 5 34 36 38 40 Flashover voltage Uc (kV) Distance a_r ( cm) dg = 2 cm, 2 rows dg = 2 cm, 3 rows dg = 2 cm, 4 rows

Fig.7.Effectofthedistancearontheinsulatorﬂashovervoltages.

3.3. Distancebetweentwosuccessiverows

Inthissectiontheeffectofthedistancearisinvestigated.The arrangement of water droplets onthesurface of thesample is designedsuchthatthereisnorowalongtheelectrodeaxis.The obtainedresultsare illustratedin Fig.7 forthree different val-uesofar.Ascanbeseentheﬂashovervoltageincreaseswiththe distancebetweentworows.Besidesthiswecanseethatthe volt-ageincreasesrapidlyuptoalimitfromwhichnoimprovementis detected.

3.4. Positionandwidthofadryzoneperpendiculartothe electrodesaxis

Fig.8shows theshape oftheinsulation performancevs the numberofrowsofwaterdropletsdepositedonthesurface perpen-diculartotheelectrodesaxis.Thetotalnumberofperpendicular rowsis7.Thenumberofparallelrowsinthiscaseisequalto5. Thedistancebetweentwosuccessiverowsisequalto1cmandthe waterdropletsconductivityisabout180␮S/cm.

ResultsofFig.8showthatwhencreatingadrybandinthe vicin-ityofthegroundedelectrode,theﬂashovervoltagepassesthrough amaximumforanairgapdistanceequaltoapproximately6cm whichcorrespondstoasinglerowofwaterdropletsincontactwith thehighvoltageelectrode.Thisoptimumcanbeexplainedbythe factthatafterdeformationofwaterdroplets(Fig.9a,b,d),the dis-chargearcoverwaterdropletsarisesfarawayfromtheendofthe highvoltageelectrodealongapathcharacterizedbyapeakinits middle(Fig.9c).

Thispathislongerthanthatbetweenthetwopointswithout waterdroplets.Thesysteminthiscaseismorerigidthanwhenthe drybandisatthehighvoltageside.However,whenthewidthof thedrybandisequaltohalfthedistancebetweentheelectrodes, thewithstandvoltageofthesystemwithadrybandinthevicinity

(5)

0 2 4 6 8 25 30 35 40 45 50

Water drops on th

e ground

side

Water drops on th

e hi

gh voltage s

ide

Flashover vo

ltage

U

C

(kV)

Distance dg (cm)

Fig.8. Relationbetweentheﬂashovervoltageandthedistancedg.

ofthehighvoltageelectrodeishigherthanthesamesystemwith thedrybandinthegroundelectrodeside.Thisdiscrepancyisdue tothefactthatdepositedwaterdropletsinthevicinityoftheHV electrodeshortentheinter-electrodepath

3.5. Positionandwidthofadrybandparalleltotheelectrodes axis

Inthis section,theeffectofthedrybanddistanceldb

paral-leltotheelectrodeaxisontheﬂashovervoltageisinvestigated. Thisinvestigationwasachievedbyremovingdropletsrowsfrom selectedareasoftheinsulatorsurface.Thewaterdropletsare char-acterizedbyanelectricconductivityequalto180␮S/cmhaving ar=1cmanddg=2cm.theinitialtotalnumberofrowsisequalto 11.

AsshowninFig.10,twoscenarioswerestudied:

a)Aftercoveringtheentiresurfaceofthesamplebywaterdroplets, rown◦6,bywhichwemeantherowalongtheelectrodeaxis,was removedﬁrst,thenrown◦5onitsrightnextcomesrown◦7on itsleft.

Thispracticecontinuesuntilonlyonerowremainsonthe sam-plesurface,

b)Inthesecondscenario,thedropletsrowswereremovedoneby onestartingfromrown◦1untilonlyonerowremainsonthe samplesurfaceaswell.

TheobtainedresultsareshowninFig.11,wherethevariation oftheﬂashovervoltageisplottedasafunctionofthedry band distanceldbpositionandwidth.

Ascanbeseen,inthecasewherewaterdropletsrowonthe elec-trodeaxiswasremoved,thedrybandcreatedcausesanincreasein thedielectricstrengthofthesystematitsmaximumvalue.From thislimitremovalofanyotherrowdoesnotproduceany improve-ment.Thismeansthattheexistenceoftherowalongtheelectrodes axiscontributessigniﬁcantlytotheshorteningofthearcingpath betweenthe two electrodes.This wasveriﬁedduring thetests byobtainingphotographicevidenceforthedischargesfollowing theinsulatorsurface(Fig.12).Itsremovaldoesnotautomatically changethepathoftheelectricdischarge,despitetheincreaseof theignitiondistanceandpartialdeformationofwaterdropletsof theadjacentrows.

Inadditiontothis,Fig.11indicatesthatadrybandhavinga widthlessthanorequalto5cmobtainedbydeletingfiverowsto theleftoftheelectrodeaxis(scenariob)doesnotaffectthesystem performanceandtheflashovervoltageisalwaysatitsminimum. However,Notonlytheremovalofthedropletsrown◦6,whichison theelectrodeaxis,increasesthedielectricstrengthofthesystem butitproducesthehighestflashovervoltageaswell.Fromthislimit noimprovementwasseenwhendeletingtherestofthedroplets rows.

4. Conclusions

Theanalysisoftheeffectofdifferenttypesofarrangementsof waterdropletsontheﬂashovervoltageofasiliconesurfacewith anon-uniformﬁeldelectrodesystemledmainlytothefollowing conclusions:

䊏Theperformanceofainsulatingsurfaceisreducedwhenitis uniformlycoveredbywaterdroplets;

䊏Undertheinﬂuenceofwaterdroplets,thereisalimitednumber ofrowsforwhichtheinsulationperformanceisminimal(about 35%lowerthanthatobtainedinthecaseofadryclean

(6)

Fig.10.Consideredscenariosofthedryband. 0 2 4 6 8 10 12 25 30 35 40 45 Flashover voltage U C (k V ) Distance l_db (cm) Scenario a Scenario b

Fig.11. Variationoftheﬂashovervoltagevsthedrybanddistanceldb.

sphere)andis quiteconstantbeyondit.Thisminimumvalue dependsonthedistancebetweentwosuccessiverowsofwater droplets;

䊏Inthecasewherewaterdropletsrowontheelectrodeaxisdoes notexist,thecreateddrybandincreasesthedielectricstrength ofthesystem;

䊏Forthecasewhereadrybandiscreatedinthevicinityofthe groundelectrodesuchthatasinglerowexistalongtheline pass-ingthroughthehighvoltageelectrode,thewithstandvoltageof thesystemishigherthanthatinthecaseofabsenceofwater dropletsonthesurface.

䊏Theﬂashovervoltagedecreasesinanon-linearmannerandis slightlyaffectedbytheincreaseofthewaterconductivity,inhigh conductivitiesregion.Finally,theeffectoftheelectrical conduc-tivityofwater dropletsonitsperformance,erectedfallingto 38%.

(7)

Fig.12.Dischargestages:caseofadrybandparalleltotheelectrodeaxis.

Acknowledgements

Thisworkisapartofa researchprojectapprovedunderthe number:J0201220120004.Theauthorswould liketothankthe MinistryofhigherEducationandScienceResearchofAlgeriafor theﬁnancialsupportofthisproject.

References

[1]R.Hackam,OutdoorHVcompositepolymericinsulators,IEEETrans.Dielectr. Electr.Insul.6(5)(1999)557–585.

[2]E.A.Cherney,R.S.Gorur,RTVsiliconerubbercoatingsforoutdoorinsulators, IEEETrans.Dielectr.Electr.Insul.6(5)(1999)605–611.

[3]J.Ndoumbe,A.Beroual,A.M.Imano,Behaviourofwaterdropletsoninsulator surfacessubmittedtoDCvoltage–coalescence,ProceedingsoftheAnnual ReportConferenceElectricalInsulationandDielectricPhenomena(CEIDP) (2012)725–728.

[4]L.Yong,X.D.Boxue,Recurrentplotanalysisofleakagecurrentindynamic droptestforhydrophobicityevaluationofsiliconerubberinsulator,IEEE Trans.PowerDeliv.28(4)(2013)1996–2003.

[5]G.G.Karady,Flashovermechanismofnon-ceramicinsulators,IEEETrans. Dielectr.Electr.Insul.6(5)(1999)718–723.

[6]M.G.Danikas,R.Sarathi,P.Ramnalis,S.L.Nalmpantis,Analysisofpolymer surfacemodiﬁcationsduetodischargesinitiatedbywaterdropletsunder highelectricﬁelds,Int.J.Electr.Electron.Eng.4(5)(2010)329–334.

[7]M.H.Nazemi,V.Hinrichsen,Partialdischargeinceptionelectricﬁeldstrength ofwaterdropletsonpolymericinsulatingsurfaces,IEEETrans.Dielectr.Electr. Insul.22(2)(2015)1088–1096.

[8]H.P.Nagaraj,N.Vasudev,K.N.Ravi,S.Aradhya,Behaviourofwaterdropletson polymersurface,ProceedingsoftheIEEE10thIntern.Conf.ontheProperties andApplicationsofDielectricMaterials(2012)1–4.

[9]Y.Zhu,K.Haji,M.Otsubo,C.Honda,N.Hayashi,Electrohydrodynamic behaviourofwaterdropletonanelectricallystressedhydrophobicsurface,J. Appl.Phys.39(2006)1970–1975.

[10]A.J.Phillips,D.J.Childs,H.M.Schneider,Agingofnon-ceramicinsulatordueto coronafromwaterdrops,IEEETrans.PowerDeliv.14(3)(1999)258–263.

[11]S.M.Rowland,F.C.Lin,Stabilityofalternatingcurrentdischargesbetween waterdropsoninsulationsurfaces,J.Appl.Phys.39(2006)3067–3076.

[12]D.A.Swift,C.Spellman,A.Haddad,Hydrophobicitytransferfromsilicone rubbertoadheringpollutantsanditseffectoninsulatorperformance,IEEE Trans.Dielectr.Electr.Insul.13(4)(2006)820–829.

[13]H.Wang,Z.Peng,S.Zhang,P.Liu,SimulationstudyonE-ﬁelddistributionand coronacharacteristicsofcompositeinsulatorwithwaterdroplets,

ProceedingsoftheAnnualReportConferenceElectricalInsulationand DielectricPhenomena(CEIDP)(2013)422–425.

[14]D.A.Swift,Flashoverofaninsulatorsurfaceinairduetopollutedwater droplets,Proceedingsofthe4thIntConf.onPropertiesandApplicationson DielectricMaterials2(1994)550–553.

[15]M.G.Danikas,P.Rakitzis,K.Karakoulidis,Studyofparametersrelatedto deteriorationphenomenaduetowaterdropletsonpolymericsurfaces,J. Electr.Eng.57(3)(2006)130–137.

[16]S.Zhu,N.Yamashita,M.Anami,C.Otsulbo,Y.Honda,Coronadischarge phenomenonandbehaviourofwaterdropletsonthesurfaceofpolymerin theACelectricﬁeld,Proceedingsofthe7thInt.Conf.onPropertiesand ApplicationsofDielectricMaterialsvol.2(2003)637–641.

[17]M.A.R.M.Fernando,S.M.Gubanski,Leakagecurrentpatternsoncontaminated polymericsurfaces,IEEETrans.Dielectr.Electr.Insul.6(5)(1999)660–667.

[18]M.G.Danikas,P.Ramnalis,R.Sarathi,Experimentalresultsonthebehaviour ofwaterdropletsonpolymericsurfacesundertheinﬂuenceofelectricﬁelds: thecaseofaninclinedtestarrangementforPVC,rubberandsiliconerubber, FunktechnikplusJ.1(2)(2013)19–39.

[19]BSEN60060-1High-voltagetesttechniques.Part1:generaldeﬁnitionsand

testrequirements,(2010).

[20]IvanJ.S.Lopes,SheshaH.Jayaram,EdwardA.Cherney,Astudyofpartial dischargesfromwaterdropletsonasiliconerubberinsulatingsurface,IEEE Trans.Dielectr.Electr.Insul.8(April(2))(2001)262–268.

[21]B.-A.L.Souza,I.J.S.Lopes,Experimentalinvestigationofcoronaonsetin contaminatedpolymersurfaces,IEEETrans.Dielectr.Electr.Insul.22(2) (2015)1321–1331.

[22]C.-A.J.Phillips,A.J.Maxwell,I.Engelbrecht,Electric-ﬁeldlimitsforthedesign ofgradingringsforcompositelineinsulators,IEEETrans.PowerDeliv.30(3) (2015)1110–1118.