Aging of Porcelain Suspension Insulators Under Mechanical and Electrical Stress on EHV AC Lines

8/16/2019 Aging of Porcelain Suspension Insulators Under Mechanical and Electrical Stress on EHV AC Lines

1/7

ESMO-95 CP-08

Agi ng of Porcel ai n Suspensi on I nsul at or sUnder Mechani cal and El ect r i cal St r ess On EHV AC Li nes

A. S J agt i ani ( Non- Member)

Seni or Engi neerAEP Ser vi ce Cor p. - Col umbus

J . R. Booker ( Member )

Seni or Engi neerAEP Servi ce Corp. Col umbus

Abst r act

A smal l number of f ai l ur es of st andard porcel ai n suspensi on i nsul ators used onAmeri can El ect r i c Power CO s AEP) 765 kV l i nes were f ound w t h t he porcel ai nshed br oken away f romt he cap and pi n assembl y, al l ow ng a l l donut l l f por cel ai nt o f al l t o the har dwar e bel ow

The i nvest i gat i on of t hese f ai l ur es was appr oached f r ommany angl es t o det erm net he sequence of event s l eadi ng t o t he f ai l ur e. Addi t i onal l y, i t was desi r ed t odet erm ne i f t hi s phenomena was t he resul t of agi ng por cel ai n under el ect r i caland/ or mechani cal st r ess.

1. Introduction

AEP's ol der 765 kV l i nes ar e desi gned usi ng 3- phase, 4- conduct or bundl esof 9 5 4 MCM 45/ 7 Rai l conduct or s st r ung i n hor i zont al conf i gur at i on w t ht wo 7#8 al um numcl ad st eel gr ound w r es suppor t ed at t he top. The newerl i nes use a bundl e of 1351.5 MCM 45/ 7 ACSR D pper conduct or s and t wo 7#8al umnum cl ad st eel gr ound w r es. The suppor t i ng st r uct ur es ar e f our -l egged sel f - suppor t i ng st eel t ower s and l at t i ce t ype guyed V al um numt ower s w t h cent er gri l l age f oundat i on and four g u y s . The t ypi cal span i s1400 t o 1500 f t . i n f l at t er r ai n.

65'+/- l 38 +/-

Typi cal Self Supporting Steel Tower Typi cal Guyed V Al um num Tower

80-7803-2506-0/95/ 4.00 01 995 IEEE


2/7

The 1 8 cent er- t o- cent er squar e bundl e of f our conduct or s i s suppor t ed bya VII st r i ng i nsul at or har dwar e assembl y usi ng ei t her si ngl e or doubl est r i ngs ( dependi ng on l oad) i n each l eg of t he 8fVrT t r i ng. The out si dephase i nsul ator hardware assembl i es use t hi r t y 1OVqx Xr1 , 25 ,000 l bs . or3 6 , 0 0 0 l bs. M&E st r engt h bal l - and- socket por cel ai n suspensi on i nsul at or s.The mddl e phase assembl y uses t hi r t y- t wo 1 f l x 5%11 nsul at or s. No corona

r i ngs are used on suspensi on assembl i es.The dead- end i nsul at or har dwar e assembl i es use f our hor i zont al st r i ngs oft hi r t y- f our 1O x 5%11 36,000 l bs. st r engt h por cel ai n bal l - and- socketsuspensi on i nsul at or s f or 1f Rai 181 onduct or , and t hi r t y- wo l l %r f x 6Ur15 0 0 0 0 l b. st r engt h f or Di pper conduct or. Both t hese dead- end hardwareassembl i es i ncl ude cor ona r i ngs.

The st udy on t he donut t ype f ai l ur e has been under way f or 2% years att hi s wr i t i ng, begi nni ng af t er t he di scover y of a br oken i nsul at or by al i ne crew dur i ng t hei r r egul ar mai nt enance i nspect i on of AEP syst emsGavi n- Mar ysvi l l e 765 kV l i ne. Al t hough t here was no i nst ance of phaseconduct or ( s) dr oppi ng t o the gr ound, t he ver y natur e of t hi s t ype off ai l ur e was i nt r i gui ng and pr ompt ed AEP's t r ansm ssi on l i ne engi neer s t oi nvest i gat e t he cause ( s ) .

To keep thi ngs i n pr oper per spect i ve, t here ar e appr oxi mat el y 2 0 0 0 ml esof 7 6 5 kV l i nes on t he AEP Syst em of whi ch appr oxi matel y 1 2 0 0 m l es havebeen i n ser vi ce f or 2 0 - 2 4 years. There ar e appr oxi matel y1, 400, 000 10f l x 3Art bal l - and- socket por cel ai n suspensi on i nsul at or s i n useon these ol der l i nes.

So f ar 2 1 i nsul at or s out of 1, 400, 000 have exhi bi t ed thi s pr obl emt o onedegr ee or another. On a percent age basi s t hi s woul d amount t o about.0015% of t he total popul at i on of i nsul at or s on t hese st r i ngs.

2 . TESTS

A test pr ogr amconsi st i ng of var i ous el ect r i cal and mechani cal t est s wasst ar t ed i n cooper at i on w t h t he i nsul at or manuf act ur er s. Bot h dest r uct i veand non- dest r uct i ve t est s were undert aken. Some cont i nue as of thi swri t i ng.

2a. The El ect r i cal Test s

Appr oxi matel y 1100 sampl e i nsul at or s of 2 5 , 0 0 0 l bs. and 36 ,000 l bs. ( M&E)r at i ng were r emoved f r omboth out si de and mddl e phases of '765 kV l i nes.Af t er t he i nsul ators were r emoved, al l were mar ked to i ndi cat e t he t owert hey wer e r emoved f r om t he dat e of t hei r r emoval , and thei r l ocat i on i nt he str i ng, w t h I11l1 ndi cat i ng t he t op uni t next t o the t ower , and 1130 ,1131 , and s 3 2 f 1 cl oser t o t he ener gi zed conduct or . Al l t he i nsul at or s wer esent t o AEP S John E. Dol an Engi neer i ng Laborat or y i n Gr ovepor t , Ohi o.

When t he i nsul at or s wer e r ecei ved at t he Dol an Labor at or y, t he i nsul at orswere vi sual l y i nspect ed f or any obvi ous manuf act ur i ng def ects or anydamage f r omhandl i ng.

Al l t he i nsul at or s wer e f l ashed over at 60 Hz t o det ect any cracks orpunct ur es i n t he porcel ai n. Thr ee hundr ed i nsul ator s wer e t hen sel ectedf or an oi l punct ur e t est per f or med as per ANSI C 2 9 . 1 St andard. Thepunct ur e vol t age was r ecor ded.

Test dat a has been gr ouped i n t he f ol l ow ng manner f or di spl ay: I nsul at or s1 t hrough 2 8 wer e t r eat ed as one l ar ge gr oup whi l e dat a f r om i nsul at ornumber s 29 , 30 , 31 , and 32 were anal yzed by posi t i on. Thi s gr oupi ng wasdone t o show t he di ver gence of t he char act er i st i cs of t he i nsul at or s neart he ener gi zed conduct or f r omt hose of t he remai ni ng uni t s.


3/7

180

170a-

-2 160

2 150

140a,

m

>"

LT

n=24

n = size of sample

I I I I30

I28 #29 #30 #31 #32

Position inString

Graph 1

Gr aph 1 shows t he oi l punct ur e vol t age of gr ouped i nsul ators . Not e t hedecr ease i n punct ur e vol t age t oward t he ener gi zed end of t he st r i ng.

2B. D ssi pat i on Factor

D ssi pat i on f act or was measur ed on sever al i nsul at or st r i ngs. Gr aph 2Ashows t he resul t i ng dat a f or one compl et e V st r i ng, demonst r at i ng ani ncr ease i n di ssi pat i on f act or at both the ener gi zed and the gr ound ends.Si m l ar measurement s were recor ded at one manuf act ur er' s l aborat ory. Forr ef erence, t he di ssi pat i on f act or was al so measur ed on ei ght i nsul at or sf r om st ock, whi ch were of si m l ar age, t ype and manuf act ur er, but hadnever been energi zed. Gr aphhese uni t s t est ed i n t he 2 . 8 t o 3. 0% r ange.2B shows t he t ypi cal change i n di ssi pat i on f act or w t ht emper at ure on one sampl e.

1 16 32

Position in StringGraph 2A

r espect t o

82


4/7

20 40 8 1M 120 140-a

TEMPERATURE n O

Graph 2B

2C. Punct ur ed I nsul at or Det ect i on ( PI D) Test

A PI D i nst r ument was used t o l ocat e punct ur ed or part i al l y cr ackedi nsul at or s on i n- ser vi ce, ener gi zed l i nes. The t ower s wher e thi s devi cewas used wer e pi cked part i al l y at r andomand par t i al l y near r oads f or easy

access. The devi ce i s at t ached t o a hot st i ck and moved f i r st down andt hen back up t he energi zed i nsul ator st r i ng whi l e t he oper ator st ands ont he t ower. As i t i s moved, an i nt er nal comput er r ecor ds t he el ect r i cf i el d magni t ude each t i me an i nsul ator shed passes. The dat a i s l aterpl ot t ed t o show t he rel at i ve f i el d magni t ude f or each i nsul at or on bot ht he downward and upward t r aver ses. The nor mal cur ve l ooks l i ke a smoothl l neckl acel l , t h a hi gh at t he energi zed end, l owt owar d the cent er and al esser hi gh at t he gr ound end of t he st r i ng. I f t her e i s a punct ur edi nsul ator , t he cur ve makes a shar p downward di scont i nui t y at t hat poi nt ,i ndi cat i ng t he change i n t he el ect r i c f i el d. Usi ng t hi s devi ce on j ustunder 5, 000 i nsul at or s, one el ect r i cal l y f ai l ed uni t was l ocat ed andr emoved and subsequent l y examned w t h t he ot her f ai l ur es.


5/7

Graph 3

Gr aph 3 shows t he curves gener at ed usi ng the P I D devi ce. Note t hedi scont i nui t y f or t he i nsul at or i n t he 17 l ocat i on. Due t o physi cali nt er f er ence, t he f i r st and l ast i nsul at or s i n t he str i ng ar e not i ncl udedi n the measurement .

2 D . X- Ray Exam nat i ons

As par t of t he i nvest i gat i on, s even i nsul at or s wer e X- r ayed. One of t heseven uni t s was previ ousl y subj ect ed to t he oi l - punct ur e t est , t wo uni t swere f i el d punct ur ed uni t s and t he r emai ni ng f our wer e bel i eved t o behi ghl y st r essed uni t s. The X- r ay showed t hat t he punct ur e pat h di d notf ol l owa st r ai ght l i ne bet ween t he met al cap on the out si de sur f ace t o t hemet al pi n on t he i nsi de surf ace of t he por cel ai n. Thi s punct ur e pat hi ndi cat es t hat t he punct ur es occur r ed over a per i od of t i me f ol l ow ng apath of l east r esi st ance whi l e crossi ng t hr ough t he por cel ai n body. Therewer e no ot her cr acks i n the por cel ai n.

MECHANICAL LOAD TESTS

Over 8 0 0 i nsul at or s of 25, 000 l bs. and 36, 000 l bs. ( M E ) st r ength wer et ested mechani cal l y t o f ai l ur e. As was done f or oi l punct ur e t est dat a,t he f ai l ur e l oad val ues of t he i nsul at or s f r om 1 t hr ough 2 8 ( posi t i on i nt he i nsul ator st r i ng) were averaged as a gr oup and the test val ues f ori nsul ator numbers 2 9 30, 31, and 32 were aver aged separat el y. The t estval ues were t hen pl ott ed on Gr aph 4 .

84


6/7

xu)

U

L

46 -

44 -

42 -

40 -

38 -

36 -

34

48

n=size of sample

46 -

44 -

42 -

40 -

38 -

36 -

34

=25=size of sample

n=26

n=33

I I

n=33

I I

I28 #29 #30 #31 #32

Position in String

Graph 4

Gr aph 4 shows a downward tr end f r omi nsul ator 29 t oward the energi zed end.

CONCLUSI ON

The f ol l ow ng concl usi ons ar e based on el ect r i cal and mechani cal t est s, X- r ayexamnat i ons and obser vat i ons on appr oxi mat el y 1100 porcel ai n suspensi oni nsul at or s over t he per i od of 2% year s. Test s were conduct ed at t womanuf actur er s' l abor atori es and at AEP's John E. Dol an Engi neer i ng Labor ator y.I t i s concl uded t hat t he punct ur e pr ocess i s t he r esul t of a combi nat i on ofsever al f actor s.

1.

2.

3 .

D el ectr i c l osses i n i nsul at i ng mat er i al s occur as a resul t of conduct i on,pol ar i zat i on and i oni zat i on. Losses i ncrease as t he t emper at ur e of t hedi el ectr i c mat er i al i ncreases. I ncreased l osses i n t ur n pr oduce f ur t heri ncr eases i n t he t emper at ur e gi vi ng r i se to the possi bi l i t y of t her malr unaway and shor t - t er mpunct ur e br eakdown. For t he i nsul ator t ypes underconsi der at i on, t hi s i s known to occur at subst ant i al l y hi gher vol t agest han have been shown t o exi st across i ndi vi dual i nsul at or s i n servi ce.

The punct ur es exam ned dur i ng thi s s t udy occur r ed i n areas of maxi mumel ect r i cal s t r ess, nor mal l y f r omt he r i mof t he met al cap t o t he pi n. Thei nsul ators were under l ower cont i nuous vol t age t han i s known t o causeshor t - t ermbr eakdown, but had been i n servi ce f or 20 year s or mor e. I t i spostul ated t hat when l ocal i zed st r ess exceeds a cert ai n val ue t hatdamagi ng di schar ges occur t hat can event ual l y resul t i n puncture[' ' .Whet her t hi s i s cont i nuous f or t he ser vi ce l i f e, or happens i n spi kesdur i ng t i mes of hi gh st r ess due t o sur f ace cont amnat i on and moi st ur e i snot known. I t woul d, however, be dependent on anomal i es i n t he i nsul atorconst r uct i on whi ch cause l ocal i zed i nt ensi f i ed st r ess such as eccent r i cpi n posi t i on or a shar p f l ashi ng on t he cap edge f r omt he cast i ng pr ocess.

Once the i nsul ator i s punct ur ed, mechani cal l oad, f r eezi ng of moi st ure anddi ggi ng due t o cont i nued arci ng appear s t o l ead t o c i r c um er e nt i al

cr acki ng. Thi s pr ocess, however , was not ver i f i ed i n t he l abor at or ytes t s .

85


7/7

Aging of Porcelain Suspension Insulators Under Mechanical and Electrical Stress on EHV AC Lines

Documents