FIST 3-30 TRANSFORMER MAINTENANCE
FACILITIES INSTRUCTIONS, STANDARDS, AND TECHNIQUES
U0+,'&*5,7,'5*&'<7R,9'0,*1F*,'*+0,'R+1R �UR'7U*1F*R'/797,+10 &'0'R*/11R7&1
F+5,*W F7/++,+'5*+05,RU/,+105 5,70&7R&5*70&*,'/0+3U'5
TRANSFORMER MAINTENANCE
1E%46*2WWW
(&R1''/,R+/*R'5'7R/*70&* ,'/0+/7*5'R+/'5*)R1U< &�HW
U0+,'&*5,7,'5*&'<7R,9'0,*1F*,'*+0,'R+1R �UR'7U*1F*R'/797,+10 &'0'R*/11R7&1
F+5,*W F7/++,+'5*+05,RU/,+105 5,70&7R&5*70&*,'/0+3U'5
TRANSFORMER MAINTENANCE
1E%46*2WWW
(&R1''/,R+/*R'5'7R/*70&* ,'/0+/7*5'R+/'5*)R1U< &�HW
U0+,'&*5,7,'5*&'<7R,9'0,*1F*,'*+0,'R+1R �UR'7U*1F*R'/797,+10 &'0'R*/11R7&1
Ac r o n y m s a n d A b b re re v i a t i o n s A
a ir
AN A
self-cooled, nonventilated
kW
kilowatt
ANSI
American National Standards
IEEE
Institu te of Electric Electrical al a nd
Institu te CEGB
Centr al Electric Electric Genera Genera ting Board
Electronic Electronic En gineers gineers M/DW
moistu moistu re by dry weight
mg
milligram
c fm
cubic feet per minute
mva
mega-volt-amps
CH 4
methane
ND
not detected
C2 H 2
acetylene
N2
nitrogen
C2 H 4
ethylene
O
oil
C2 H 6
ethane
O2
oxygen
CO
carbon m onoxide onoxide
OD
outer diameter
CO 2
carbon dioxide dioxide
ppb
parts per billion
CT
current tran sfo sformer
ppm
par ts per millio million n
DBPC
Ditertiary But yl Par acresol acresol
psi
pounds per squar e inch inch
D GA
dissolved gas analysis
R e c l a m a t i o n Bureau of Reclamation
EH V
extra high voltage voltage
SCADA
FA
forced forced air (fan (fan s)
FO
forced oil (pumps)
G
some type of gas
GA
gas, self-cooled
TDCG
tota l dissolved dissolved combu combu stible gas
gm
grams
TOA TO A
Tran sformer sformer Oil Analyst Analyst
GS U
generator generator step u p
TTR TT R
tran sfo sformer t urns ratio test
H2
hydrogen
TS C
Technical Service Cent er
ID
inner diameter
UV
ultraviolet
IF T
interfacial tension
V
volts
IE C
Int erna tiona tiona l Electrotechnic Electrotechnical al
W
water /oil oil hea t exchan exchan ger
Commission IR
infrared
J HA
job job hazard an alysis alysis
KOH
potassium hydroxide hydroxide
kV
kilovolt
k VA
kilovoltampere
Supervisory Supervisory Control and Dat a Acqu Acqu isition
S TP
standard temperature and pressur e
Contents Page
1 . Pu P u r p os e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . L a s t u p d a t e d O ct ob e r 2 0 0 0
1
2 . I n t r od od u ct ct io ion t o R ec ecl a m a t io i on T r a n sf sfor m e r s . . . . . . . . . . . La La s t u p da da t e d O ct ct ob ob er er 20 20 00 00
1
3 . Tr T r a ns n s fo for m er er C ooli n g M e t h od od s . . . . . . . . . . . . . . . . . . . . . L a st s t u pd p d a t ed ed O ct ob ob e r 2 0 0 0 3 .1 D r y T yp e T r a n s for m e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 3.1.1 Potent ial Pr oblems oblems a nd Rem edial Actions Actions for for Dr y Type T r a n s for m e r C ooli n g S y s t e m s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 .2 L iq u i d -I m m e r s e d T r a n s for m e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . 2 . 1 L i q u i d -I m m e r s e d , Ai r -C ool e d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . 2 . 2 L i q u id id -I -I m m er er s ed ed , Ai r -C -C oo ool e d /F /F or or ce ce d L i q u id id -C -C oo ool e d . . . . . . . . . . . . . . . . . 3 . 2 . 3 L i q u i d -I m m e r s e d , W a t e r -C ool e d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . 2 . 4 L i q u id id -I m me m e r se se d , F or or ce ce d L iq u id id -C oole d . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.5 3.2.5 Potent ial Pr oblems oblems a nd Rem edial Actions Actions for for Liquid F illed illed T r a n s for m e r C ooli n g S y s t e m s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . 2 . 6 C ooli n g S y s t e m I n s p e ct i on s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 3
4 . O i l-F il il l e d T r a n sf s for m er e r I n sp sp ec ect io ion s . . . . . . . . . . . . . . . . . La La s t u p da da t ed e d O ct ct ob ob e r 2 0 0 0 4 .1 O i l-F il le d T r a n s for m e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 1 . 1 T r a n s for m e r T a n k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 1 . 2 T op O il T h e r m om e t e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 1 . 3 Wi Wi n d i n g T e m p e r a tu t u r e T h e r m om om et et e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 1 . 4 O i l L e v e l I n d i ca t or s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 1 . 5 P r e s s u r e R e l ie f D e vi ce s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 .1 .6 S u d d e n P r e s s u r e R e la y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 1 . 7 B u ch h olz R e la y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.8 4.1.8 Tran sform sform er Bush ings: Testing and Maint ena nce of H ig h -Vol t a ge B u s h in g s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 .2 O i l P r e s e r v a t i on S e a li n g S y s t e m s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 .2 .1 S e a lin g S ys t e m s T yp e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 2 . 2 G a s P r e s s u r e C on t r ol C om p on e n t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 .3 G a s k e t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 .4 T r a n s for m e r O il s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 4 . 1 T r a n s for m e r O i l F u n ct i on s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 4 . 2 Di Di s s ol ve d G a s An a l ys i s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 4 . 3 K e y G a s M e t h od . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.4 4.4.4 Tran sform sform er Diagn osis osis Using In dividua dividua l an d Tota Tota l Dissol Dissolved ved K e y G a s C on ce n t r a t i on s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 4 . 5 R oge r s R a t i o M e t h od of D G A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 .5 M oi s t u r e P r ob le m s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 5 . 1 D i s s olv e d M oi s t u r e i n T r a n s for m er er O i l . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 5 . 2 Mo M ois t u r e in T r a n s for m e r I n s u la t i on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Tra Tra nsformer Oil Tests Tha t Sh ould Be Done Done Annu ally With With t he Dissolved Dissolved G a s An a l y s i s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 6 . 1 Di Die le ct r i c S t r e n gt h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
3 6 6 7 7 7 8 9 10 11 11 11 11 13 14 16 17 18 21 21 23 28 35 35 35 37 37 48 53 57 58 61 61
C o n t e n t s ( c o n t .) Page
4.6.2 In t er fa cia l Ten sion (IF T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 4.6.3 Acid Nu m ber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 4.6.4 Test for Oxygen In h ibit or . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 4.6.5 P ower F a ct or . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 4.6.6 F u r a n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 4.6.7 Ta kin g Oil Sa m ples for DGA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 4.6.8 Silicon e Oil-F illed Tr a nsfor m er s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 4.7 Tr a n sfor m er Test in g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 4.7.1 Win din g Resist a n ces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 4 .7 .2 C or e I n su la t ion R es is t a n ce a n d I n a dve r t en t C or e G r ou n d T es t . . . . . . . . 7 7 Refer en ces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
80
Tables Ta ble N o.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Page
Dr y Type Tr an sfor mer Ma in ten an ce Su mm ar y . . . . . . . . . . . . . . . . . . . . . . . . Oil-F illed Tr an sfor mer Ma in ten an ce Su mm ar y . . . . . . . . . . . . . . . . . . . . . . . . Tr a nsfor m er Ga sket Applica t ion Su m m ar y . . . . . . . . . . . . . . . . . . . . . . . . . . . Ver t ica l G roove C om p res sion for C ir cu la r N it r ile Ga s ket s . . . . . . . . . . . . . . . Ve r t ica l G r oove C om p r es sion for R ect a n gu la r N it r ile G a sk et s . . . . . . . . . . . . D is s ol ve d K ey G a s C on ce n t r a t i on L im i t s i n P a r t s P e r M ill ion (p p m ) . . . . . . . Act ion s Ba sed on Dissolved Com bu st ible Ga s . . . . . . . . . . . . . . . . . . . . . . . . . TO A L1 Lim it s a n d Ge ner a t ion R at e P er M on t h Ala r m Lim it s . . . . . . . . . . . F a u lt Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dissolved Ga s Solu bilit y in Tr an sfor mer Oil . . . . . . . . . . . . . . . . . . . . . . . . . . Roger s Ra t ios for Key Ga ses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Typica l F a u lt s in P ower Tr an sfor m er s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Com pa r ison of Wa ter Dist ribu tion in Oil a nd P a per . . . . . . . . . . . . . . . . . . . . Doble Lim it s for In -Ser vice Oils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Addit ion a l Gu idelin es for In -Ser vice Oils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Com pa r ison of Ga s Lim it s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Su ggest ed Levels of Con cer n (Lim it s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Doble and IEEE Physical Test Limits for Service-Aged Silicone Fluid . . . . . . Tr a n sfor m er Test Su m m a r y Ch a r t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 5 31 31 33 38 39 44 45 47 50 54 55 64 65 73 73 75 79
Figures Figure N o.
1 2
Page
Typica l Oil F low . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oil Level In dica t or . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
6 13
C o n t e n t s ( c o n t .) F i g u r e s ( c o n t .) Figure N o.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Page
P r essu r e Relief Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Su dden P r essu r e Rela y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bu ch h olz Rela y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F r ee Br ea t h in g Tr a n sfor m er . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P r essu r ized Br ea t h in g Tr a n sfor m er . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P r essu r ized In er t Ga s Tr a n sfor m er . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ga s P r essu r e Con t r ol Com pon en t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F r ee Br ea t h in g Con ser va t or . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Con ser va t or wit h Bla dder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bla dder F a ilu r e Rela y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Au xilia r y Sea lin g Syst em . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cr oss Sect ion of Cir cu la r Ga sk et in Gr oove . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross Section of Gasket Remains Constan t Before Tightening and After . . . . Bowin g a t F la n ges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bolt Tigh t en in g Sequ en ces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Com bu st ible Ga s Gen er at ion Ver su s Tem per at ur e . . . . . . . . . . . . . . . . . . . . . Maximu m Amoun t of Wat er Dissolved in Minera l Oil Ver su s Tem per a t u r e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tr an sfor mer Oil P er cen t Sa tu r at ion Cu rves . . . . . . . . . . . . . . . . . . . . . . . . . . Wa ter Dist ribu tion in Tr an sfor mer In su la tion . . . . . . . . . . . . . . . . . . . . . . . . . Myer s Mu lt iplier Ver su s Tem per a tu r e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wa t er Con t en t of P a per a n d Oil Nom ogr a m . . . . . . . . . . . . . . . . . . . . . . . . . . . In ter fa cia l Ten sion , Acid N um ber , Yea rs in S er vice . . . . . . . . . . . . . . . . . . . . Oil Sa m plin g P ipin g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sa m ple Syr in ge (F lu sh in g) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sa m ple Syr in ge (F illin g) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sa m ple Syr in ge Bu bble Rem ova l . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relationship of Oxygen to Car bon Dioxide an d Car bon Monoxide a s Tr a n sfor m er Ages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vii
14 16 17 21 22 23 23 25 26 26 27 32 34 35 36 41 56 57 58 59 60 62 68 68 69 69 71
1. PURPOSE
This docum ent is to provide guidance to Burea u of Reclam at ion (Reclam at ion) powerplan t personnel in ma intena nce, diagnostics, and testing of tr an sform ers a nd a ssociated equipment.
2. INTROD UCTION TO RECLAMATION TRANS FORMER S
Tra nsformer s ra ted 500 kilovoltam peres (kVA) and a bove are considered power tr ansform ers. Reclama tion ha s hun dreds of power tra nsform ers with voltages as low as 480 volts (V) an d a s h igh a s 550 kilovolts (kV). All genera tor st ep-up (GSU) tra nsform ers, an d m an y sta tion service, and excitation tr ansform ers a re considered power tr an sform ers becau se they a re r at ed 500 kVA or lar ger. Sta nda rds organ izat ions such as American Nat iona l Sta nda rds Institu te/Institute of Electrical and Electronic Engineers (ANSI/IEEE) consider average GSU transformer life to be 20 to 25 year s. This estimate is based on continu ous operation at ra ted load a nd service conditions with a n a verage ambient t empera tu re of 40 °C (104 °F) and a temper at ure r ise of 65 °C. This estimate is also based on th e assum ption th at t ra nsform ers receive adequate ma inten an ce over th eir service life [24]. Reclam at ion, Bonn eville Power Administr at ion, an d Western Area P ower Administra tion condu ct r egular stu dies to determ ine stat istical equipment life. These studies show that average life of a Reclama tion t ra nsform er is 40 years. Reclama tion gets longer service tha n IEE E estima tes becau se we operat e at lower am bient t empera tu res an d with lower loads. A significan t nu mber of tr an sform ers were pur cha sed in th e 1940s, 1950s, and into the 1970s. Several have been replaced, but we have man y that ar e near ing, or a re alrea dy well past, th eir ant icipated service life. We should expect tra nsform er r eplacement an d failures to increase due t o this age factor. Curr ent m inimum r eplacement time is around 14 mont hs; a more realistic time ma y be 18 month s to 2 years. In th e fut ur e, lead times ma y extend well beyond wha t th ey are today. Therefore, high quality maintena nce and accura te diagnostics ar e importa nt for a ll tr an sform ers, but a bsolutely essentia l for older ones—especially for critical tr an sform ers th at would cau se loss of genera tion. It is also very import an t to consider providing spar es for critical transformers.
3. TRANS FORMER COOLING METHODS
Hea t is one of th e most common destr oyers of tr an sform ers. Opera tion at only 10 °C above the t ra nsform er r ating will cut tr an sform er life by 50%. Heat is cau sed by intern al losses due to loading, high ambient tempera tur e, an d solar ra diation. It is import an t to un der stan d how your par ticular tr an sform ers a re cooled and h ow to detect pr oblems in the cooling system s. ANSI a nd IE EE r equire th e cooling class of each tr an sform er to appea r on its na mepla te. Cooling classifications, with short explan at ions, appea r in sections 3.1 and 3.2. The letter s of th e class designat e inside atm ospher e an d type or types of cooling. In some tra nsform ers, more th an one class of cooling and load ra ting ar e indicat ed. At each
1
DRY TYPE TRANSFORMER MAINTENANCE SUMMARY See Section 3.1 When new after energizing and allowing temperature and loading to stabilize
Do an infrared scan and compare with temperature gage, if any. If transformer is gas filled (nitrogen [N2]), check pressure gage against data sheets; never allow gas pressure to fall below 1 pound per square inch (psi). Check loading and compare with nameplate rating. Functionally test fans and controls for proper operation. Functionally test temperature alarms and annunciator points. Check area around transformer clear of debris and parts storage. Check transformer room for proper ventilation.
After 1 week of operation at normal loading
Perform infrared scan and compare with temperature gage, if any. Check temperature gage, if any, and compare with nameplate rating. Check loading and compare with nameplate rating.
Annually
Perform an infrared scan before de-energizing.
(Note: The time between these periodic inspections may be increased if the first internal inspection of windings and connections are found clean and in good condition and if loading is at or below nameplate rating.)
De-energize and remove panels for internal inspection. Use vacuum to remove as much dirt as possible. After vacuuming, use low pressure dry air (20 to 25 psi) to blow off remaining dirt. Caution: Make sure air is dry. Check for discolored copper and discolored insulation. Check for corroded and loose connections. Check for carbon tracking on insulation and insulators. Check for cracked, chipped, and loose insulators. If windings are found dirty, add filter material to air intake ports. Check fan blades for cleanliness; remove dirt and dust. Check fans, controls, alarms and annunciator points. Check pressure gage on N2 filled transformers; compare with weekly data sheets; never allow pressure to fall below 1 psi. Repair all problems found in above inspections.
2
step of add itiona l cooling, th e ra ting increa ses to corr espond with increa sed cooling. Note that the letter “A” indicates air, “FA” indicates forced air (fans), “O” indicates oil, “FO” indicates forced oil (pumps), “G” indicates some type of gas, and “W” indicates there is a water /oil heat excha nger. 3.1 Dry Type Transform ers
Cooling classes of dry type tr an sform ers a re covered by ANSI/IEE E st an dar d C57.12.01 Section 5.1 [1]. A sh ort expla na tion of ea ch cla ss is given below. 1. Class AA ar e ventilated, self-cooled tr ansform ers. This means th at t here ar e ventilat ion ports locat ed in outside walls of the tr an sform er enclosur e. There a re n o fans t o force air into an d out of th e enclosur e with typically no extern al fins or ra diators. Cooler air enter s the lower port s, is heat ed as it rises past windings, an d exits the up per vent ilation port s. (It will not be repeat ed below; but it is obvious tha t in every cooling class, some hea t is also rem oved by nat ur al circulat ion of air ar ound th e outside of th e enclosur e.) 2. Clas s AFA tr a ns form er s ar e self-cooled (A) an d add itiona lly cooled by forced circulation of air (FA). This means tha t t here a re ventilation ports for fan inlets a nd out lets only. (Inlets are usua lly filtered.) Norm ally, th ere ar e no add itiona l ventilat ion ports for na tur al a ir circulation. 3. Class AA/FA tr an sform ers a re vent ilated, self-cooled (sam e as Cla ss AA in item 1). In a ddition, th ey have a fan or fan s providing additional forced-air cooling. Fa ns m ay be wired to sta rt a ut oma tically when the tem perat ur e reaches a pre-set value. These tr an sform ers gener ally have a dua l load r at ing, one for AA (self-cooling na tu ra l air flow) and a larger load rating for FA (forced air flow). 4. Class ANV tr an sform ers a re self-cooled (A), non-ventilat ed (NV) un its. The enclosur e ha s no vent ilation port s or fan s an d is not sealed to exclude m igrat ion of out side air, but t here ar e no provisions t o inten tiona lly allow outside air t o enter a nd exit. Cooling is by na tu ra l circulat ion of air ar ound the enclosur e. This tra nsform er ma y have some type of fins a tt ached outside th e enclosur e to increa se sur face ar ea for additional cooling. 5. Class GA tr an sform ers a re sea led with a gas inside (G) an d ar e self-cooled (A). The enclosur e is her met ically sealed to preven t leaka ge. These tra nsform ers typically ha ve a gas , such as n itr ogen or freon, to provide high dielectr ic an d good heat rem oval. Cooling occur s by na tu ra l circulat ion of air a roun d the outside of th e enclosure. Ther e ar e no fan s to circulat e cooling air; however, ther e ma y be fins a tt ached t o th e out side to a id in cooling. 3 .1 .1 P o t e n t i a l P r o b l e m s a n d R e m e d i a l A c t i o n s f o r D r y T y p e Tr a n s f o r m e r C o o l i n g S y s t e m s [ 14 ] . It is import an t t o keep tra nsform er enclosures reasonably clean. It is also importa nt to keep th e area ar ound th em clear. Any items nea r or against t he tr an sform er impede heat t ra nsfer to cooling air ar oun d
3
th e enclosure. As dirt a ccum ula tes on cooling sur faces, it becomes m ore a nd m ore difficult for a ir ar oun d the t ra nsform er to remove heat. As a r esult, over time, the transformer temperature slowly rises unnoticed, reducing service life. Tran sform er rooms a nd vaults sh ould be ventilated. Porta ble fan s (never water ) ma y be used for a dditional cooling if necessar y. A fan r at ed at a bout 100 cubic feet per min ut e (cfm) per kilowat t (kW) of tr an sform er loss [5], locat ed nea r t he top of th e room t o remove hot air, will suffice. These r ooms/vault s sh ould not be used as st ora ge. Wh e n t h e t r a n s f o r m e r i s n e w , check t he fan s an d all cont rols for pr oper operat ion. After it has been energized and the loading and temper at ur e are sta ble, check the tem perat ur e with an infra red (IR) cam era a nd compa re loading with the na meplate. Repeat the tem perat ure checks after 1 week of operat ion. O n c e e a c h y e a r un der normal load, check tr an sform er tempera tur es with an IR cam era [4,7]. If the temper at ur e rise (above ambient) is near or above na meplate r at ing, check for overloading. Check the tempera tur e alarm for pr oper operat ion. Check enclosur es and vau lts/rooms for dirt a ccum ula tion on tr an sform er sur faces and debris near or against en closur es. Remove all items nea r en ough to affect air circulat ion. To avoid dust clouds , a vacuum should first be used to remove excess dirt. Low pressu re (20 to 25 pounds per squ ar e inch [psi]) dry compr essed air m ay be used for cleaning after most dirt ha s been removed by vacuu m. The tr an sform er m ust be de-energized before t his procedure un less it is totally enclosed an d th ere a re n o exposed ener gized condu ctors. Porta ble genera tors ma y be used for lighting.
After de-energizing the transformer, remove access panels and inspect windings for dirt - an d heat -discolored insu lat ion an d str uctu re problems [14]. It is importa nt tha t dirt not be allowed to accum ulat e on windings because it impedes hea t rem oval an d redu ces winding life. A vacuum s hould be used for th e initia l winding clean ing, followed by compr essed air [7]. Car e mu st be ta ken t o ensu re th e compressed a ir is dry to avoid blowing moistu re int o windings. Air pr essur e should not be great er t ha n 20 to 25 psi to avoid imbedding sma ll par ticles into insu lat ion. After clean ing, look for discolored copper an d insu lat ion, which indicat es overh ea tin g. If discolora tion is foun d, check for loose conn ections. If th ere a re n o loose conn ections, check t he cooling pat hs very carefully an d check for overloading after the tr an sform er ha s been re-ener gized. Look for car bon tr acking an d cracked, chipped, or loose insu lators. Look for an d r epair loose clam ps, coil spa cers, deteriorat ed bar riers , and corroded or loose conn ections. Check fans for proper operation including controls, temperature switches, and ala rm s. Clean fan blades and filter s if needed. A dirt y fan blade or filter redu ces cooling air flow over th e windings a nd r educes service life. If vent ilation ports do not ha ve filters, th ey ma y be fabricated from home-fur na ce filter ma ter ial. Adding filters is only necessary if th e windings ar e dirty u pon yearly inspections.
4
OIL-FILLED TRANSFORMER MAINTENANCE SUMMARY Task
After 1 Month of Service
Annually
Oil pumps load current, oil flow indicators, fans, etc. See 3.2.5, 3.2.6, and 4.1. Thermometers 4.1.2 and 3. Heat exchangers. Transformer tank 4.1.1. Oil level gages 4.1.4. Pressure relief 4.1.5. Do a DGA.
Oil pumps load current, oil flow indicators, fans etc, see 3.2.5, 3.2.6 and 4.1 Thermometers 4.1.2 and 3, heat exchangers Transformer tank 4.1.1 Oil level gages 4.1.4 Pressure relief 4.1.5 Do a DGA
IR scan of transformer cooling system, bushings and all wiring.
See 3.2.5 and 4.1.8.
See 3.2.5 and 4.1.8.
Test all controls, relays, gages; test alarms and annunciator points.
See 3.2.5, 4.1.4, 4.1.5.
See 3.2.5 Inspect pressure relief for leaks and indication for operation (rod extension) see 4.1.5
Inspect transformer bushings.
Check with binoculars for cracks and chips; look for oil leaks and check oil levels. IR scan. See 4.1.8.
check with binoculars for cracks and chips, look carefully for oil leaks and check oil levels IR Scan See 4.1.8
3 to 5 Years
Before energizing, inspect and test all controls, wiring, fans alarms, and gages. Indepth inspection of transformer and cooling system, check for leaks and proper operation. Do a DGA.
Indepth inspection of bushings, cleaning waxing if needed.
Check diaphragm or bladder for leaks if you have conservator. See 4.2.2.
Thermometers. See 4.1.3. Oil level gages 4.1.4. Inspect pressure relief 4.1.5. Sudden pressure relay 4.1.6. Buchholz relay 4.1.7. Test alarms, fan and pump controls, etc. See 3.2.6.
Close physical inspection, cleaning/ waxing, and Doble testing, plus checks in boxes above left. See 4.1.8.
Doble test transformer and bushings.
Doble test transformer and bushings before energizing. See 4.1.8, 4.7.
Inspect pressure controls if you have a nitrogen over oil transformer. Inspect pressure gage.
See 4.2.2.
See 4.1.8 and 4.7.
See 4.2.2. Also see 4.2.1 to test pressure gage if trans has N2 over oil with no means to automatically add N2.
5
Check pr essure gages by looking at th e weekly data sheets; if pressur e never var ies with tem pera tu re chan ges, th e gage is defective. Never allow the press ur e to go below about 1 psi during cold weat her. Add nitrogen t o bring the pr essure u p to 2½ to 3 psi to insur e th at moist a ir will not be pulled in.
3.2 Liquid-Imme rsed Transforme rs
Cooling classes of liquid-immer sed t ra nsform ers a re covered by IEE E C57.12.00 Section 5.1 [2]. A sh ort expla na tion of ea ch cla ss follows: 3.2.1 Liquid -Imme rse d, Air-Coole d. There a re t hree classes in t his category.
1. Class OA: Oil-imm ersed, self-cooled. Tra nsform er windings an d core ar e imm ersed in some t ype of oil and ar e self-cooled by na tu ra l circulat ion of air ar oun d the outside enclosure. Fins or radiat ors m ay be atta ched to the enclosur e to a id in cooling. 2. Clas s OA/FA: Liquid-imm er sed, self-cooled/forced a ir-cooled. Sa m e as OA above, with the addition of fan s. Fa ns ar e usually mounted on radiators. The tr an sform er t ypically has t wo load r at ings, one with t he fans off (OA) an d a lar ger ra ting with fans opera ting (FA). Fa ns ma y be wired to start a utoma tically at a pre-set temperat ure. 3. Clas s OA/FA/FA: Liqu id- imm ers ed, self-cooled/forced air-cooled/forced air-cooled. Sa me a s OA/FA above with an add itiona l set of fans. Ther e typically will be th ree load r at ings corr esponding to each increment of cooling. Increased ra tings are obtained by increasing cooling air over port ions of the cooling su r faces. Typically, there ar e radiators att ached to th e ta nk to aid in cooling. The t wo groups of fans m ay be wired to star t au tomat ically at pre-set levels as temperature increas es. Ther e ar e no oil pum ps. Oil flow th rough the tr an sform er windings is by th e nat ura l principle of convection (heat rising).
Figu re 1.—Typical Oil Flow .
6
3.2.2 Liquid-Imme rsed , Air-Cooled/Forced Liquid-Cooled. There are two classes in this group.
1. Clas s OA/FA/FOA: Liquid-imm er sed, s elf-cooled/for ced air -cooled/forced liquid, an d forced air-cooled. Windings an d core a re imm ersed in some type of oil. This tr an sform er typically ha s radiat ors at ta ched to th e enclosure. The tra nsform er has self-cooling (OA) natural ventilation, forced air-cooling FA (fans), and forced oil-cooling (pu mp s) with ad ditiona l forced air -cooling (F OA) (more fan s). The tr an sform er ha s thr ee load r at ings corr esponding to each cooling step. Fa ns an d pum ps may be wired to sta rt a utomat ically at pr e-set levels as temper at ur e increases. 2. Clas s OA/FOA/FOA: Liquid-imm er sed, self-cooled/forced oil, an d forced aircooled/forced oil, an d forced air-cooled. Cooling contr ols ar e ar ra nged t o st ar t only par t of the oil pum ps an d par t of the fans for t he first load r at ing/temper atu re increase, and th e rema ining pum ps and fans for th e second load ra ting increase. The na meplate will show at least th ree load r at ings. 3.2.3 Liqu id-Imme rse d, Wate r-Coole d. This category has t wo class es.
1. Class OW: Tra nsform er coil an d core ar e imm ersed in oil. Typically a oil/wat er heat excha nger (radiator) is at ta ched to the outside of th e tan k. Cooling water is pum ped th rough th e hea t exchanger, but th e oil flows only by natu ra l circulation. As oil is hea ted by th e windings, it r ises to th e top and exits th rough piping to the ra diator. As it is cooled, the oil descends th rough the r adiator a nd r e-enters the tra nsformer t ank at t he bottom. 2. Class OW/A: Tra nsform er coil an d core ar e imm ersed in oil. This tra nsform er ha s two rat ings. Cooling for one r at ing (OW) is obta ined as in 1 a bove. The selfcooled r at ing (A) is obtained by nat ur al circula tion of air over t he t an k a nd cooling surfaces. 3.2.4 Liquid-Imme rsed , Force d Liquid-Cooled. This cat egory ha s two classes.
1. Clas s FO A: Liquid-imm ers ed, forced liquid-cooled with forced air-cooled. This tr an sform er norma lly has only one ra ting. The tr an sform er is cooled by pumping oil (forced oil) th rough a r adia tor norm ally att ached to th e out side of th e ta nk . Also, air is forced by fan s over t he cooling su rfa ce. 2. Class FOW: Liquid-immersed, forced liquid-cooled, water cooled. This tr an sform er is cooled by an oil/water heat excha nger n orm ally moun ted separ at ely from the ta nk. Both t he tra nsform er oil an d the cooling water a re pum ped (forced) thr ough th e hea t exchan ger t o accomplish cooling.
7
3 .2 .5 P o t e n t i a l P r o b l e m s a n d R e m e d i a l A c t i o n s f o r L i qu i d F i l l e d T ra n s f o r m e r C o o l i n g S y s t e m s . L e a k s . Tan ks a nd r adia tors m ay develop oil leaks, especially at conn ections. To repair a leak in a r adiator core, you mu st rem ove the r adiator. Sma ll leaks may also develop in head ers or individua l pipes. These sm all leak s possibly ma y be stopped by peening with a ba ll peen ha mm er. Some m an ufactu rer’s field personnel try t o stop leaks by using a two-par t epoxy while th e tr an sform er is un der vacuum . Do not try this unless the tr an sform er has been drained, becau se a vacuum ma y cau se bubbles to form in th e oil th at can lodge in t he winding and cause a rcing. When a ll else fails, the leak m ay be welded with oil still in t he ra diat or, if proper pr ecaut ions a re car efully observed [3, 4]. Welding with oil inside will cause gases t o form in th e oil. Take a n oil sam ple for a dissolved gas an alysis (DGA) before welding a nd 24 h our s a fter r e-ener gizing to identify gas increases du e to welding. If the leak is bad enough, the t an k ma y have to be dra ined so the leak can be repaired. Treat leaks carefully; do not ignore them . Oil leaks ar e serious m ainten an ce and en vironment al issues and sh ould be corr ected. Radiators ma y need to be cleaned in ar eas where deposits a ppear on pipes and header s. Dirt and deposits hamper hea t tra nsfer to the cooling air. Finned ra diators mu st be cleaned with compr essed air when th ey become dirt y. P l u g s . A ft e r 1 m o n t h o f s e r v i c e a n d y e a r l y , p e r f o r m a n I R s c a n a n d p h y s i c a l i n s p e c t i o n o f r a d i a t o r s a n d t r a n s f o r m e r t a n k s [4,7]. Partially plugged radia tors will be cooler th an t hose perform ing norma lly. You m ay also feel the ra diator pipes by hand. Plugged radiator sections or individual pipes/plenums will be noticeably cooler; however, you will not be able to reach all of th em. Radia tors m ay become plugged with sludge or foreign debr is; th is usu ally occurs in water tu bes on th e oil/wat er hea t excha nger. Do not forget to check t he bleed line for t wo-walled hea t exchan gers.
If plugged r adia tors a re d iscovered, t hey n eed to be corrected a s soon a s possible. Some ra diators are at ta ched to the main ta nk with flan ges an d have isolating valves. These ma y be removed for clean ing and/or leak repa ir without dr ainin g oil from the tr an sform er. If ra diators are at ta ched directly to the m ain ta nk, oil mu st be drained before clean ing th em. If ra diat ors are plugged with sludge, chan ces ar e the tr an sform er is sludged up also. In th is case, th e oil should be repr ocessed an d the tr an sform er cleaned interna lly. Competent cont ra ctors should be obta ined if th is is necessar y. S l u d g e . If tempera tu re seems to be slowly increasing while the t ra nsform er is operat ing under the sam e load, check t he DGA for moistu re, oxygen, and t he inter facial ten sion (IFT). The combina tion of oxygen a nd m oistu re causes sludging, which m ay be revea led by a low IFT nu mber . Sludge will slowly build up on windings and core, an d th e temper at ur e will increase over time.
8
V a l v e P r o b l e m s . If your tr an sform er h as isolating valves for ra diators, check to ma ke sur e they ar e fully open on both t op and bottom of th e rad iators. A broken valve stem m ay cause t he valve to be fully or pa rt ially closed, but it will app ear th at th e valve is open. M i n e r a l D e p o s i t s . Don’t even t hink a bout spraying water on t he r adiators or ta nk t o increase cooling except in th e most dire emergency. Minera ls in th e water will deposit on r adiators a s wat er evaporates a nd a re a lmost impossible to rem ove. These m inera ls will redu ce the efficiency of cooling still fur th er. Additiona l fan s blowing on r adiat ors a nd/or t ra nsform er t an k is a better alternative [4].
One IR scan performed on a t ran sformer run ning at higher tha n n orma l temper atu re r evealed tha t t he oil level was below the u pper ra diator inlet pipe, which prevent ed oil circulat ion. The oil level indicator was d efective an d stu ck on norma l. These indicat ors m ust be tested a s ment ioned below. 3 .2 .6 C o o l i n g S y s t e m In s p e c t i o n s . A ft e r 1 m o n t h o f s e r v i c e a n d y e a r l y , i n s p e c t a n d t e s t t h e f a n s . Look at th e fan s anytime you a re ar oun d tr an sform ers in the switchyar d or in the powerplant . If it is a hot day and tr an sform ers ar e loaded, all the fans should be ru nning. If a fan is stopped and th e rest of the group is runn ing, the ina ctive fan should be repaired. During an inspection, th e temper atu re cont roller should be adjusted t o start all the fans. Listen for un usua l noises from fan bearings an d loose blades an d repa ir or replace fau lty fan s. Bad bearings can a lso be detected with an IR scan if the fan s ar e running. A ft e r 1 m o n t h o f s e r v i c e a n d y e a r l y , i n s p e c t a n d t e s t t h e o i l p u m p s . Inspect piping and connections for leaks. Override the temperature controller so th at t he pump sta rt s. Check the oil pum p motor cur rent on all thr ee pha ses with an accur at e am met er; th is will give an indicat ion if oil flow is correct a nd if un usu al wear is cau sing additional motor loadin g. Record th is inform at ion for lat er compa rison, especially if ther e is n o oil flow indicator. I f t h e m o t o r l o a d c u r r e n t i s l o w , s o m e t h i n g i s c a u s i n g l o w o i l fl o w . Car efully inspect all valves to ma ke sure t hey are fully open. A valve stem m ay break a nd leave the valve par tially or fully closed, even th ough th e valve han dle indicat es th e valve is fully open. Pu mp im pellers ha ve been foun d loose on t he sh aft, redu cing oil flow. Sludge bu ildup or debr is in lines can a lso cau se low oil flow. If moto r load c u r r e n t i s h i g h , t h i s m a y i n d i c a t e i m p e d e d p u m p r o t a t i o n . Listen for un usua l noises. Thru st bearing wear r esults in th e impeller advan cing on th e housing. An impeller t ouching the h ousing ma kes a ru bbing sound which is different from th e soun d of a failing motor bear ing. If th is is hea rd, rem ove the pum p motor from t he housing and check impeller cleara nce. Replace the th ru st bearing if needed, and replace the m otor bearings if the sh aft ha s too much play or if noise is un usu al.
9
Three pha se pumps will run a nd pum p some oil even when t hey are ru nning backwar ds. Vane type oil-flow met ers will indicate flow on th is low am oun t. The best indication of th is is tha t sometim es the pu mp will be very noisy. The motor load curr ent m ay also be lower th an for full load. If th is is suspected due t o th e extra n oise and higher tr ansform er tem perat ur e, the pum p should be checked for proper rota tion. Reverse two pha se leads if th is is encoun ter ed.[4] A ft e r 1 m o n t h o f s e r v i c e a n d y e a r l y , c h e c k t h e o i l f lo w i n d i c a t o r . I t h a s a sma ll paddle which extends int o the oil strea m a nd m ay be either on th e suction or discha rge side of the pu mp . A low flow of only about 5 feet p er s econd velocity causes t he flag to rotat e. Flow can be too low, and t he indicator will still show flow. If ther e is no flow, a sprin g retu rn s th e flag to the off position an d a swit ch provides an a larm . With cont rol power on th e switch, open t he pu mp circuit a t t he motor sta rter and m ake sure t he correct alarm point activates when the pum p stops. Check th at the point er is in th e right position when t he pum p is off an d when it is run ning. Pointer s can stick an d fail to provide an alar m when n eeded. Oil flow ma y also be checked with a n u ltr asonic flow met er. Ultr asonic listenin g devices can detect worn bearings, rubbing impellers, an d other unu sual n oises from oil pumps . P u m p s c a n p u l l a i r in t h r o u g h g a s k e t s o n t h e s u c t i o n s i d e o f t h e p u m p s . The suction (vacuum ) on t he inta ke side of the pum p can pull air t hr ough gaskets th at a re not tight. Pu mp suction ha s also been known to pull air th rough packing ar oun d valve stems, in th e suction side piping. This can result in da ngerous bubbles in th e tr an sform er oil and ma y cau se the gas detector or Buchholz relay to operat e. Dissolved gas a na lysis will show a big increa se in oxygen a nd n itr ogen cont ent [4]. High oxygen and n itrogen conten t can a lso be cau sed by gasket leaks elsewhere. A ft e r 1 m o n t h o f s e r v i c e a n d y e a r l y , i n s p e c t w a t e r -o i l h e a t e x c h a n g e r s . Test an d inspect t he pum ps as m entioned above. Look for a nd r epair leaks in piping an d heat excha nger body. Exam ine the latest dissolved gas ana lysis results for dissolved moistur e and free water. If free wat er is present an d ther e ar e no gasket leaks, the water port ion of the wa ter-oil heat excha nger mu st be pressur e tested. A leak m ay ha ve developed, allowing water to migra te into the tr an sform er oil, which can destr oy the tr an sform er. If the h eat excha nges piping is double-walled, check th e dra in for wa ter or oil; check m an ufactu rer ’s inst ru ction manual.
4. OIL-FILLED TRANSF ORMER INSP ECTIONS
A tr an sform er ma intena nce progra m mu st be based on t h o r o u g h routine inspections. These inspections m ust be in addition to normal da ily/weekly data gather ing trips t o check oil levels an d temper at ur es. Some monitoring ma y be done rem otely usin g super visory cont rol and da ta acquisition (SCADA) system s, but t his can n ever substitute for thorough inspections by competent maintenance or operations people. 10
4.1 Oil-Filled Transform ers Af t e r 1 m o n th o f s e r v i c e a n d o n c e e a c h y e a r , m a k e a n i n d e p t h i n s p e c t i o n o f oil-filled transformers. Before beginnin g, look carefully at t emper at ur e an d oil level data sheets. If temperat ur e, pressure, or oil level gages never cha nge, even with seasonal tempera tu re an d loading cha nges, someth ing is wrong. The gage ma y be stu ck or da ta sheet s ma y have been filled in incorr ectly. Exam ine th e DGA’s for evidence of leak s, et c. 4.1.1 Transforme r Tank. Check for excessive corrosion a nd oil leaks. Pa y special att ention to flanges an d gaskets (bushings, valves, and ra diators) and lower section of the m ain t an k. Report oil leaks t o maintena nce, an d pay special at ten tion to the oil level indicat or if leaks a re foun d. Severely corr oded spots should be wire brush ed and pa inted with a ru st inhibitor. 4.1.2 Top Oil Thermo me ters. These a re t ypically sealed spiral-bour don-tube dial indicat ors with liquid-filled bulb sensors. The bulb is norma lly inside a th ermometer well, which penet ra tes th e tan k wall into oil near th e top of the ta nk. As oil temper at ur e increases in the bulb, liquid expands, which expands th e spiral tube. The tube is at ta ched to a pointer tha t indicat es temper at ur e. These pointers ma y also have electrical cont acts t o trigger a larm s a nd sta rt cooling fans as tem perat ur e increases. An extra point er, norm ally red, indicat es maximum temper atu re since the last t ime the indicat or was reset. This red pointer rises with th e main pointer but will not decrease unless ma nua lly reset; thu s, it a lways indicat es the highest tempera tu re reached since being set. See the instru ction manual on your specific transformer for details. 4 .1 .3 Wi n d i n g T e m p e r a t u r e T h e r m o m e t e r s . These devices ar e supposed to indicat e hottest spot in the winding based on th e man ufactur ers heat ru n tests. At best, t his device is only accur at e at top nam eplate r at ed load a nd t hen only if it is not out of calibra tion [17]. They are not wha t th eir na me implies and can be misleadin g. They ar e only winding h o t t e s t - s p o t s i m u l a t o r s an d not very accur at e. There is no temper at ur e sensor imbedded in the winding hot spot. At best, they provide only a rough approximation of hot spot winding temperature an d should not be relied on for a ccur acy. They can be used to tur n on additional cooling or a ctivat e alar ms a s th e top oil th erm omet ers do.
Winding tempera tu re ther mometers work t he sam e as the top oil th ermometer (4.1.2) above, except t ha t t he bulb is in a separa te t herm ometer well near the t op of the tan k. A wire-type heat er coil is eith er inserted int o or wr apped ar oun d the th ermometer well which sur rounds the tem perat ur e sensitive bulb. In some tr an sform ers, a curr ent t ra nsform er (CT) is ar ound one of the t hr ee winding leads and provides current directly to the heater coil in proportion to winding current. In other tra nsform ers, the CT supplies cur rent t o an a ut o-tr an sform er tha t supplies curr ent to the heater coil. The heater wa rm s the bulb an d the dial indicat es a tempera tu re, but it is not the tru e hottest -spot temper at ur e.
11
These devices ar e calibrat ed at the factory by cha nging taps either on t he CT or on th e au totran sform er, or by adjusting th e calibration r esistors in th e contr ol cabinet. They norm ally can not be field calibrat ed or t ested, other tha n t esting the th ermometer, a s m entioned. The calibrat ion r esistors can be a djusted in t he field if the m an ufactur er pr ovides calibration curves for t he tr an sform er. In pr actice, most winding temper at ur e indicat ors a re out of calibration, an d th eir readings ar e mea ningless. These temper at ure indicat ions sh ould not be relied upon for loading operations or maintenance decisions. Fiber optic temper at ur e sensors can be imbedded directly int o the winding as the tr an sform er is being built and a re mu ch m ore accura te. This system is available as a n option on new t ra nsform ers a t a n increased cost, which m ay be wort h it since th e tr ue winding “hottest-spot” temperat ure is critical when higher loading is required. Therm ometer s can be r emoved without lowering the t ra nsform er oil if they a re in a th ermometer well. Check your t ra nsform er instru ction ma nua l. Look car efully at the capillary tubing between th e ther mometer well an d dial indicat or. If the tu bing has been pinched or accidently stru ck, it may be restr icted. This is not an obvious defect, and it can cause t he dia l pointer to lock in one position. If this defect is found, th e whole gage must be retu rn ed to the factory for r epair or repla cement ; it can not be repa ired in the field. Look for a leak in th e tubing system ; th e gage will be readin g very low an d mu st be r eplaced if a lea k is discovered. Therm ometer s should be removed an d tested every 3 to 5 year s as described below. Thermometer Testing. Every 3 to 5 yea rs, and if trouble is sus pecte d, do a t h e r m o m e t e r t e s t i n g . Suspend th e indicat or bulb and an a ccur at e mercury th erm ometer in a n oil bath . Do not allow eith er to touch the side or bottom of th e cont ainer. Heat the oil on a h otplate while stirring and compa re th e two th ermometers while the temper at ure increases. If a magnet ic stirr ing/heat ing plate is available, it is more effective th an h an d stirring. Pa y particular at tent ion to th e upper tempera tur e ran ge at which your t ra nsform ers normally operat e (50 °C to 80 °C). An ohm met er should also be used to check switch opera tions. If either dial indicat or is more tha n 5 °C different th an th e mercury th ermometer, it should be replaced with a spa re. A nu mber of spar es should be kept, based on t he quan tity of tr an sform ers at t he plant. Oil bath t est kits are available from the Qua litrol Compan y. After calling for Qu alitr ol au th orizat ion a t 716-586-1515, you can ship defective dial th erm ometer s for rep air a nd calibrat ion to: Qua litrol Co., 1387 Fair port Rd., Fair port, NY 14450.
The a lar ms a nd other functions sh ould also be tested t o see if th e correct an nu nciat or points a ctivate, pum ps/fan s opera te, etc. If it is not possible to replace the t emper at ur e gage or sen d it to th e factory for repair, place a tem perat ur e corr ection factor on your data form to add to the dial reading so the corr ect t empera tur e will be recorded. Also lower th e alar m a nd 12
pum p-tur n-on sett ings by this sam e corr ection factor. Since these a re pressu refilled systems, t he ind icat or will typically rea d low if it is out of calibra tion. Field test ing has sh own some of th ese gages readin g 15 °C to 20 °C lower th an a ctua l temper atu re. This is ha zardous for t ra nsform ers becau se it will allow them t o cont inuously run hotter t ha n int ended, due to delayed alar ms a nd cooling activat ion. If th ermometers a re not tested and err ors corr ected, tran sform er service life may be shortened or premature failure may occur. 4.1.4 Oil Leve l Indicat ors. After 1 mon th of serv ice , insp ec t and every 3 t o 5 y e a r s , c h e c k t h e t a n k o i l l e v e l i n d i c a t o r s . These ar e float operat ed, with th e float mechanism ma gnetically coupled thr ough t he t an k wall to the dial indicat or. As level increases, the float r ota tes a ma gnet inside the ta nk. Outside the tan k, another ma gnet follows (rotat es), which moves th e pointer. The center of th e dial is norma lly mar ked with a temperatu re 25 °C (77 °F). High a nd low level point s ar e also marked to follow level cha nges as t he oil expands and cont ra cts with temperature chan ges. Th e proper Figu re 2.—Oil Leve l Indicato r. way to determ ine accura te oil level is to first look at t he t op oil tem pera tu re indicator. After determ ining the tempera tu re, look at th e level gage. The pointer should be at a reasonable level corr esponding to the t op oil tempera tur e. If the tr an sform er is fully loaded, t he t op oil temp era tu re will be high, an d th e level indicat or should be near the high ma rk. If th e tra nsform er is de-energized and th e top oil temper at ur e is near 25 °C, the oil level pointer should be at or n ear 25 °C.
To check t he level indicator, you can rem ove th e out side mecha nism for t esting without lowering tr an sform er oil. After r emoving the gage, hold a magnet on th e back of the dial and r ota te th e ma gnet; the dial indicat or sh ould also rotate. If it fails to respond or if it dr ags or st icks, rep lace it. As ment ioned a bove, defective un its can be sent t o the factory for r epair. Ther e ma y also be electr ical switches for a lar ms a nd possibly tr ipping off th e tr an sform er on falling tan k level. These should be checked with an ohmm eter for proper operat ion. The alar m/tr ipping circuits should also be tested to see if th e corr ect an nu nciat or points a nd relays respond. See the tra nsform er instru ction book for information on your specific indicator.
13
If oil ha s ha d to be lowered in t he tr an sform er or conser vat or for other rea sons (e.g., inspections), check th e oil level float mecha nism . Rotat e th e float mecha nism by han d to check for free movement . Check the float visually to mak e sure it is secure to the arm and th at t he arm is in the proper shape. Some arm s ar e form ed (not st ra ight). 4 .1 .5 P r e s s u r e R e l i e f D e v i c e s . These devices ar e th e tr an sform ers’ last line of defense against excessive inter na l press ur e. In case of a fault or short circuit, th e resulta nt ar c insta ntly vaporizes surrounding oil, cau sing a ra pid buildup of gaseous pressu re. I f t h e p r e s s u r e r e l i e f d e v i c e d o e s n o t o p e r a t e p r o p e r l y a n d p r e s s u r e i s n o t s u f fi c i e n t l y r e l i e v e d w i t h i n a f e w m i l l i s e c o n d s , a catastrophic tan k rupture can res ult, spreading flaming oil over a wide a r e a . Two types of th ese devices ar e discussed below. The inst ru ction m an ua l for your tr an sform er m ust be consulted for specifics. C a u t i o n : Never paint pressur e-relief devices becau se paint can cau se th e plunger or rotat ing shaft to stick. Then t he device might not relieve pressu re, which could lead to cat ast rophic ta nk failur e dur ing a fau lt. Look at th e top of th e device; on newer u nits , a yellow or blue but ton should be visible. If th ese ha ve been pa inted, th e button will be the same color a s th e tan k. On older un its, a red flag should be visible; if it ha s been pa inted, it will be th e sam e color as t he t an k. If th ey ha ve been pain ted, they should be repla ced. It is virt ua lly impossible to rem ove all paint from th e mechan ism a nd be certa in th e device will work when needed. COLORED ROD SHOWS PROTECTIVE COVER New er Pressure Relief TRIPPED POSITION Devices. Newer pressure COMPRESSION SPRINGS relief devices a re spr ing-loaded valves tha t a utomat ically ALARM SWITCH reclose following a press ur e release. The springs are held in compression by t he cover an d press on a disc which sea ls DIAPHRAGM an opening in the ta nk top. If pressur e in th e tan k exceeds operat ing pressure, the disk GASKETS moves upwar d a nd r elieves SWITCH RESET TRANSFORMER TANK pressur e. As pressur e LEVER decreases, the springs reclose th e valve. After opera ting, th is F i g u r e 3 .—P r e s s u r e R e l i e f D e v i c e . device leaves a brightly colored r od (brigh t yellow for oil, blue for s ilicone) exposed a ppr oxima tely 2 inch es a bove th e top. This rod is easily seen u pon inspection, alt hough it is not always visible from floor level. The r od ma y be reset by pressing on t he t op un til it is aga in recessed into the device. The switch m ust also be ma nu ally reset. A relief device is shown in t he open position in figure 3 a bove.
14
C a u t i o n : Bolts th at hold the device to the ta nk m ay be loosened safely, but never loosen screws wh ich h old th e cover t o th e flan ge with out referr ing to the instru ction man ual and using great car e. Springs th at oppose ta nk pressur e ar e held in compression by these screws, and their stored energy could be hazardous. O n c e e a c h y e a r , an d as soon a s possible after a known thr ough-fau lt or intern al fault, inspect pressu re devices to see if th ey ha ve opera ted. This mu st be done from a h igh-lift bucket if th e tra nsform er is energized. Look at ea ch pressur e r elief device to see if th e yellow (or blue) but ton is visible. If th e device ha s operat ed, about 2 inches of th e colored r od will be visible. Ea ch year, test th e alar m circuits by operat ing the switch by han d an d ma king sure th e corr ect an nu nciator point is activat ed. If th e relief device opera tes dur ing opera tion, do not re-energize the tra nsform er; Doble and other testing m ay be required before re-ener gizing, and a n oil sam ple should be sent for a na lysis E v e r y 3 t o 5 y e a r s , when doing other m ainten an ce or t esting, if the tr an sform er ha s a conservat or, exam ine the top of th e tra nsform er ta nk a round th e pressure relief device. If oil is visible, th e device is leaking, either ar ound t he t an k gask et or relief diaph ra gm. If th e device is 30 year s old, repla ce th e whole un it. A nitr ogen blan keted t ra nsform er will use a lot m ore nitr ogen if th e relief device is leakin g; th ey should be tested a s described below.
A test sta nd with a pr essure gage may be fabricated to test the pressur e relief function. Cur ren t cost of a pr essur e relief device is about $600, so test ing inst ead of replacement ma y be prudent . Ha ve a spar e on ha nd so th at t he ta nk will not ha ve to be left open. If th e ta nk t op or pressu re r elief device ha s gasket limiting grooves, always u se a nit rile replacemen t gask et; if th ere a re n o grooves, use a cork -nitrile gask et. Relief devices th emselves do not leak often ; th e gasket u sua lly leaks. O l d e r P r e s s u r e R e l i e f D e v i c e s. Older pressur e relief devices h ave a diaphra gm an d a r elief pin tha t is destroyed each time th e device operat es and m ust be replaced. C a u t i o n : Th ese par ts m u s t b e replaced with exact r eplacement par ts, or t he operat ing r elief-pressur e of th e device will be wrong.
The relief pin determ ines operat ing pressure; a n um ber, which is th e operat ing pres sur e, norm ally appea rs on top of th e pin. Check your specific tr an sform er instru ction m an ual for proper cat alog numbers. Do not assum e you ha ve the right par ts, or tha t corr ect pa rt s ha ve been pr eviously insta lled—look it u p. If the operat ing pressur e is too high, a cata strophic tank failure could resu lt. On older un its, a shaft rotates, operat es alarm /tr ip switches, and r aises a sma ll red flag when the un it releases pressur e. If units ha ve been painted or ar e more th an 30 years old, they should be replaced with the new m odel as soon a s it is possible to have a tra nsform er outa ge. 15
Once each year a nd a s soon a s possible after a thr ough-fau lt or intern al fault, exam ine the indicator flag to see if th e device ha s opera ted. They mus t be exam ined from a high-lift bucket if th e tra nsform er is energized. A clear an ce mu st be obtained to test, repair, or reset t he device. See the instru ction ma nua l for your sp ecific tr an sform er. Test ala rm /tr ip circuits by opera ting th e switch byhand. Check to ma ke sure th e corr ect an nu nciat or point a ctivates. E v e r y 3 t o 5 y e a r s , when doing other m ainten an ce or t esting, examine th e top of th e tran sform er tan k around th e pressure relief device. If th e tran sform er has a conser vat or and oil is visible, th e device is leak ing, eith er a roun d th e ta nk ga sket or relief diaphra gm. The gasket and/or device must be replaced. Take care tha t th e new device will fit th e sam e tan k opening prior t o ordering. Most of th em a re ma de by the Qua litrol Compa ny; cont act th e ma nufacturer to obtain a corr ect replacement. 4 .1 .6 S u d d e n P r e s s u r e R e l a y . Int ern al a rcing in an oil-filled power tr an sform er can insta ntly vaporize surrounding oil, genera ting gas pressur es tha t can cau se cat ast rophic failure, rupt ur e the t an k, and spread flaming oil over a large area . This can da ma ge or destr oy oth er equipment in addition to the tra nsformer and presents extreme hazards t o workers.
The relay is designed to detect a sudden pressur e increase caused by arcing. It is set t o operat e before the pr essur e relief device. The cont rol circuit should deenergize the tr an sform er an d provide an a larm . The relay will ignore norma l pressur e cha nges such as oil-pum p sur ges, temper at ur e cha nges, etc. Modern su dden p ressu re r elays consist of thr ee bellows (see figure 4) with silicone sealed inside. Chan ges in pressure in th e tra nsform er deflect th e main sensing bellows. Silicone inside a cts on t wo cont rol bellows ar ra nged like a ba lan ce beam , one on each side. One bellows senses pr essur e cha nges th rough a sm all orifice. The opening is automatically changed by a bimetallic strip to adjust for normal temperature CONTROL ORIFICE BIMETAL TEMPERATURE COMPENSATOR changes of the oil. CONTROL BELLOWS BLEED VALVE The orifice delays PRESSURE BALANCE BEAM pressur e cha nges in SILICONE FLUID ACTUATOR ELECTRICAL SWITCH th is bellows. The SENSING BELLOWS other bellows responds to TO TRANSFORMER immediate pressure cha nges and is affected mu ch more TRANSFORMER OIL quickly. Pressure difference tilts th e ELECTRICAL CONNECTOR balance beam and 1/8" DIAMETER DRAIN HOLE activat es th e switch. HOUSING SILICONE SENSING FLUID MANIFOLD CONTROL BELLOWS This type relay au tomat ically resets F i g u r e 4 .—S u d d e n P r e s s u r e R e l a y . 16
when t he two bellows again rea ch pr essure equilibrium. If th is relay operat es, do not re-energize the tr an sform er un til you h ave determined t he exact cau se and corrected the problem. Old style sudden pr essur e relays ha ve only one bellows. A sudd en excessive pressur e with in the t ra nsform er ta nk exerts pr essure directly on t he bellows, which moves a spring-loaded operat ing pin. The pin opera tes a switch which provides alar m an d breaker tr ip. After th e relay ha s operat ed, the cap must be removed and th e switch reset t o norm al by depressing the reset butt on. O n c e e v e r y 3 t o 5 y e a r s , the su dden pressu re r elay should be tested a ccording to ma nu factur er’s inst ru ctions. Gener ally, only a squeeze-bulb an d pres sur e gage (5 psi) ar e required. Disconnect th e tripping circuit a nd u se an ohmm eter t o test for r elay operat ion. Test the a larm circuit an d verify that the corr ect a larm point is activated. Use an ohm met er to verify the tr ip signa l is activated or, if possible, apply only contr ol voltage to the breaker an d ma ke sur e th e tr ipping function operat es. Consu lt the ma nu factu rer ’s man ua l for your specific tr an sform er for detailed instr uctions. 4 .1 .7 B u c h h o l z R e l a y (f o u n d o n l y o n t r a n s f o r m e r s w i t h c o n s e r v a t o r s ) . The Buchh olz relay has t wo oil-filled cham bers with floats a nd r elays ar ra nged vertically one over th e oth er. If high eddy cur ren ts, local overh eat ing, or part ial discha rges occur with in th e ta nk, bubbles of resultan t gas rise to th e top of the ta nk. These rise through the pipe between the tank a nd th e conservat or. As gas bubbles migrat e along th e pipe, th ey enter t he Buchholz relay and rise into the top cham ber. As gas builds up inside the cham ber, it displaces the oil, decreasin g th e level. The t op float descends with oil level un til it passes a ma gnetic switch which activat es an a larm . The bottom float a nd relay can not be activated by additiona l gas buildup. The float is locat ed slight ly below th e top of the pipe so th at once the t op cham ber is filled, add itiona l gas goes int o th e pipe and on up to th e conservat or. Typically, inspection windows are provided so tha t th e amount of gas an d relay opera tion ma y be viewed during test ing. If th e oil level falls low enough (conser vat or empt y), switch conta cts in th e bott om chamber ar e activated by the bottom F i g u r e 5 .—B u c h h o l z R e l a y .
17
float. These conta cts are typically conn ected to cause th e tra nsform er to trip. This relay also serves a third function, similar to the sudden pressu re relay. A ma gnetically held paddle a tt ached t o the bottom float is positioned in th e oil-flow str eam between the conservat or an d tran sform er tan k. Normal flows resulting from tempera tur e chan ges ar e small and bypass below the paddle. If a fault occur s in th e tra nsform er, a pressu re wave (sur ge) is creat ed in the oil. This surge tr avels thr ough th e pipe and displaces th e paddle. The paddle activat es the sam e ma gnetic switch as t he bottom float ment ioned a bove, tripping th e tr an sform er. The flow rat e at which th e paddle activates th e relay is norm ally adjusta ble. See your specific tran sform er instru ction ma nua l for details. O n c e e v e r y 3 t o 5 y e a r s while the transformer is de-energized, functionally test th e Buchhholz relay by pumping a sm all amount of air into the t op chamber with a squeeze bulb hand pum p. Watch the float operation through the window. Check to make sur e the corr ect alar m point h as been activated. Open the bleed valve an d vent air from t he chamber. The bottom float a nd switching can not be tested with air pr essure. On some relays, a r od is provided so th at you can t est both bott om an d top sections by push ing the floats down u nt il the tr ip points a re activat ed. If possible, verify th at t he brea ker will tr ip with th is operat ion. A voltohm meter ma y also be used to check the switches. If these conta cts a ctivate dur ing operat ion, it m eans tha t t he oil level is very low, or a pressur e wave ha s activat ed (bott om cont acts), or t he tr an sform er is gassing (top conta cts). If th is relay operat es, do not re-energize the t ra nsform er u ntil you h ave determ ined the exact cau se. 4.1.8 Transforme r Bush ings : Testi ng and Mainte nan ce of High-Voltage B u s h i n g s . When bush ings are n ew, they should be Doble tested as a n a cceptan ce test . Refer to the M4000 Doble test s et inst ru ctions, th e Doble Bush ing Field Test Guide [8], and t he ma nu factu rer ’s dat a for guida nce on acceptable resu lts. C a u t i o n : Do not t est a bush ing while it’s in its wood shipping cra te, or while it is lying on wood. Wood is not a s good an insu lator a s porcelain a nd will cau se th e readings to be inaccur ate. Keep the test resu lts as a baseline record to compa re with futur e tests. A ft e r 1 m o n t h o f s e r v i c e a n d y e a r l y , check th e extern al porcelain for cra cks an d/or cont am inat ion (requir es binoculars ). Ther e is no “perfect insulat or”; a sma ll am ount of leakage cur rent always exists. This cur rent “leaks” thr ough an d along the bush ing surface from th e high-volta ge conductor to ground. If th e bushing is damaged or heavily contaminated, leakage current becomes excessive, an d visible evidence ma y appear as car bon t ra cking (tr eeing) on t he bush ing sur face. Fla shovers ma y occur if th e bushings ar e not cleaned per iodically.
Look car efully for oil leaks . Che ck th e bush ing oil level by viewing th e oil-sight glass or th e oil level gage. When t he bush ing has a ga ge with a pointer , look carefully, becau se the oil level should vary a litt le with tem pera tu re chan ges. If th e pointer n ever changes, even with wide am bient t empera tur e and load chan ges, 18
th e gage should be checked at t he next out age. A stu ck gage point er coupled with a sm all oil leak can cause explosive failur e of a bush ing, dama ging the tr an sform er an d oth er switchyar d equipment. A costly extended outa ge is the result. If the oil level is low an d th ere is a n ext ern al oil leak, check th e bolts for pr oper torque an d th e gasket for pr oper compr ession. If torque a nd compr ession a re corr ect, the bushing mu st be replaced with a spar e. Follow instr uctions in the tr an sform er ma nua l car efully. It is very import an t tha t the corr ect type gasket be inst alled and t he corr ect compr ession be applied. A leaky gasket is pr obably also leaking wat er an d air into the tr an sform er, so check the most recent t ra nsform er DGA for h igh moistur e an d oxygen. If the oil level is low an d th ere is no visible externa l leak, t here m ay be an intern al leak ar ound the lower seal into the tr an sform er ta nk. If possible, re-fill the bushing with t he sa me oil and car efully monitor th e level and t he volume it t akes to fill the bushing to th e proper level. If it tak es more than one quar t, mak e plan s to replace the bush ing. The bushing mu st be sent t o the factory for r epair or it mu st be junked; it can not be repaired in t he field. C a u t i o n : Never open th e fill plug of an y bushin g if it is at an elevated temper atu re. Some bush ings have a nitr ogen blan ket on top of th e oil, which pres sur izes as th e oil expan ds. Always consu lt the ma nu factu rer ’s inst ru ction ma nua l which will give the tem perat ur e ran ge at which th e bushing may be safely opened . Gen er ally, th is will be betwee n 15 °C (59 °F ) an d 35 °C (95 °F). Pr essur ized hot oil ma y sudden ly gush from t he fill plug if it is r emoved while at elevated temper at ur e, cau sing burn h azar ds. Generally, the bush ing will be a little cooler tha n t he top oil temper at ur e, so this tem perat ur e gage may be used as a guide if the gage ha s been tested a s men tioned in 4.1.3.
About 90% of all preventable bushing failures are caused by moisture entering th rough leaky gaskets, cra cks, or seals. Int erna l moisture can be detected by Doble testin g. See F IST 3-2 [9] and Doble Bushing F ield Test Guide [8] for tr oubles and corr ective actions. Int erna l moistur e cau ses deteriora tion of the insu lat ion of th e bush ing and can resu lt in explosive failur e, cau sing extensive tr an sform er an d oth er equipment dam age, as well as ha zards to worker s. A ft e r 1 m o n t h o f s e r v i c e a n d y e a r l y , examine the bushings with a n IR cam era [4,7]; if one ph ase sh ows a ma rkedly higher temper atu re, th ere is probably a bad conn ection. The conn ection a t t he top is usua lly th e poor one; however, a bad conn ection inside the t ra nsform er ta nk will usua lly show a h igher t empera tu re at th e top as well. In a ddition, a bad conn ection inside the tr an sform er will usu ally show hot m etal gases (ethan e an d ethylene) in th e DGA. O n c e e v e r y 3 t o 5 y e a r s , a close ph ysical ins pection a nd clean ing should be done [9]. Check car efully for leak s, cra cks, an d car bon tr acking. This inspection will be required more often in a tm ospheres where salts a nd dust deposits appear on th e
19
bush ings. In condit ions t ha t pr oduce deposits, a light ap plicat ion of Dow Corn ing grea se DC-5 or GE In sulgel will help redu ce risk of extern al flash over. The downside of th is treatm ent is tha t a grease buildup may occur . In high hum idity an d wet a reas, a better choice ma y be a h igh quality silicone pa ste wa x applied to th e porcelain, which will redu ce th e risk of flash over. A spr ay-on wa x cont ainin g silicone, such a s Tur tle Wax bra nd, h as been foun d to be very u seful for cleanin g an d waxing in one opera tion, providing th e deposits ar e not too ha rd. Wax will cause water t o form beads rat her t ha n a continu ous sheet, which r educes flash over risk. Cleanin g ma y involve just spra ying with Tur tle Wax and wiping with a soft cloth . A lime r emoval product, such a s “Lime Away,” also ma y be useful. More stubborn contaminates may require solvents, steel wool, and brush es. A high pressure water st ream ma y be required to rem ove salt an d other wat er soluble deposits. Limestone powder blasting with dr y air will safely remove meta llic oxides, chemicals, salt-cake, an d almost a ny ha rd conta mina te. Other materials, such as potters clay, walnut or pecan shells, or crushed coconut shells, ar e also used for ha rd cont am inat es. Car bon dioxide (CO 2 ) pellet blast ing is more expensive but virtu ally elimina tes cleanu p becau se it evaporat es. Ground up corn -cob blast ing will remove soft polluta nt s su ch a s old coatings of built-up grea se. A comp eten t experienced cont ra ctor sh ould be employed an d a th orough written job hazar d ana lysis (J HA) perform ed when an y of these tr eatm ents a re used. Corona (air ionizat ion) ma y be visible at tops of bush ings at twilight or night, especially durin g periods of ra in, mist, fog, or h igh hu midity. At the t op, corona is considered norm al; however, as a bushing becomes m ore a nd more conta minat ed, corona will creep lower an d lower. If the bus hing is not clean ed, flash over will occur when corona n ear s the grounded tr an sform er top. If corona seem s to be lower th an th e top of the bush ing, inspect, Doble test, a nd clean the bushing as quickly as possible. If flash over occurs (pha se to ground fau lt), it could dest roy th e bushin g and cause a n extend ed out age. Line-to-line fau lts also can occur if all th e bushings ar e conta mina ted an d flash over occurs. A corona scope may be used to view an d ph otogra ph low levels of corona indoors u nder norm al illumina tion an d outdoors a t t wilight or n ight. High levels of corona ma y possibly be viewed outdoors in th e daytime if a da rk background is a vailable, such a s tr ees, can yon walls, buildings, etc. The corona scope design is prima rily for indoor a nd n ight tim e use; it can not be used with blue or cloudy sky background . This techn ology is ava ilable at t he Techn ical Ser vice Cent er (TSC), D-8450. C a u t i o n : See the t ra nsform er ma nua l for deta iled instru ctions on cleaning and repa iring your sp ecific bush ing surfaces. Differen t solvents , wiping ma ter ials, and clean ing met hods may be requ ired for differen t bush ings. Differen t repa ir techn iques ma y also be required for sm all cra cks an d chips. Gener ally, glypta l or insulating varn ish will repair sm all scra tches, ha irline cracks, and chips. Sha rp edges of a chip should be honed sm ooth, an d th e defective ar ea pa inted with insu lat ing var nish t o provide a glossy finish. Ha irline cra cks in th e surface of th e porcelain m ust be sealed becau se accum ulated dirt an d moistu re in th e cra ck ma y resu lt in flash over. Epoxy should be used to repa ir lar ger chips. If a bush ing
20
insulator ha s a lar ge chip tha t r educes the flashover distan ce or h as a large cra ck totally thr ough the insulat or, the bushing must be replaced. Some ma nufacturer s offer r epair service to dam aged bushings th at can not be repa ired in th e field. Conta ct t he m an ufactur er for your par ticular bushings if you h ave repair questions. O n c e e v e r y 3 t o 5 y e a r s , depending on the atmosphere and service conditions, th e bushin gs should be Doble test ed. Refer to Doble M-4000 test set inst ru ctions, Doble Bushin g Field Test Guide [8], FIST 3-2, [9] and t he m an ufactu rer ’s instru ctions for pr oper values a nd t est pr ocedures. Bushings should be cleaned prior to Doble test ing. Cont am ina tion on th e insu lat ing surface will cause th e resu lts to be ina ccur at e. Testing ma y also be done before and after cleanin g to check met hods of clean ing. As the bush ings age an d begin to deter iora te, redu ce th e testing interval to 1 year. Keep accur at e records of results so tha t repla cement s can be order ed in advan ce, before you h ave to remove bushings from service.
4 .2 O i l P r e s e r v a t i o n S e a l i n g S y s t e m s
The pur pose of sealing systems is t o prevent air a nd m oisture from cont am inating oil an d cellulose insulat ion. Sealing systems a re designed to preven t oil inside the tr an sform er from comin g into cont act with air. Air cont ain s moistu re, which cau ses sludging and an a bun dan t supply of oxygen. Oxygen in combina tion with moistur e causes grea tly accelerat ed deter iora tion of th e cellulose. This oxygen-moistur e combina tion will great ly reduce service life of th e tr an sform er. Sealing systems on m an y existing Reclama tion power tr an sform ers a re of the iner t gas (nitrogen) pressure design; however, we have ma ny oth er designs. Curr ent pra ctice is to buy only conservat or designs with bla dder s for tra nsformer voltages 115 kV an d above an d capa cities a bove 10 m ega-voltam ps (mva ). Below these values, we buy only inert gas pr essure system tra nsform ers, as depicted in figur e 8. Some of the sealing systems ar e explained below. Ther e ma y be var iations of each design, an d not every design is covered. The order below is roughly from earlier to more modern. 4 .2 .1 S e a l i n g S y s t e m s T y p e s . F r e e Br e a t h i n g . Sealing systems have progressed from early designs of “free breat hing” ta nk s, in which an air space on top of the oil is
F i g u r e 6 .—F r e e B r e a t h i n g Transformer.
21
vented to atmosphere thr ough a br eath er pipe. The pipe typically is screened to keep out insects and r odents a nd tu rn ed down to prevent r ain from ent ering. Breat hing is cau sed by expansion an d contr action of th e oil as temper at ur e cha nges. These earlier designs did not use an air dryer, and condensa tion from moistur e formed on inside walls and ta nk t op. Moistu re, oxygen, and n itr ogen would also dissolve directly int o oil from t he a ir. This was n ot the best des ign. As men tioned before, a combinat ion of oxygen an d moistur e accelera tes det eriorat ion of cellulose insu lat ion. Moistu re also decreases dielectr ic str ength , destr oying insu lat ing quality of th e oil, an d causes form at ion of sludge. If you ha ve one or more of these earlier design t ra nsform ers, it is recomm ended th at a desiccant t ype air dr yer be added to the breat her pipe. S e a l e d o r P r e s su r i z e d B r e a t h i n g . This design is similar to the free breat hing one with add ition of a pressur e/vacuu m bleeder valve. When the tr an sform er was installed, pressur ized dry air or nitr ogen wa s placed on top of th e oil. The bleeder valve is designed to hold pressure inside to approximately plus or minus 5 psi (figure7). The F i g u r e 7 . —P r e s s u r i z e d B r e a t h i n g sam e problems with Transformer. moistur e a nd oxygen occur as pr eviously described. Pr oblems ar e not as severe becau se “brea th ing” is limited by the bleeder valve. Air or N 2 is exha usted to the outside atmosphere when a positive pressur e more tha n 5 psi occurs inside the ta nk . This process does not add m oistu re a nd oxygen to th e tan k. However, when cooling, the oil cont ra cts an d, if pressu re falls 5 psi below th e out side at mospher e, the valve allows out side air int o the ta nk, which pu lls in moisture a nd oxygen. O n c e e a c h y e a r , check the pr essure gage against th e weekly data sheets; if the pressur e never varies with seasonal tem perat ure cha nges, the gage is defective. Add nitrogen if th e pressu re falls below 1 psi to keep m oistu re lad en a ir from being pulled in. Add enough N 2 to bring the pressu re t o 2 to 3 psi. P r e s s u r i z e d I n e r t G a s S e a l e d S y s t e m . This system keeps spa ce above the oil pres sur ized with a dr y iner t gas, norma lly nitr ogen (figure 8). This design prevents a ir an d moistur e from coming into cont act with insulat ing oil. Pr essure is ma intained by a nitr ogen gas bott le with the pressur e regulated norm ally between 0.5 and 5 psi. Pr essur e gages are provided in th e nitr ogen cubicle for both h igh and low pressu res (figure 9). A press ur e/ vacuum gage is norma lly
22
conn ected t o rea d low pressu re gas inside the ta nk. This gage ma y be locat ed on th e tra nsform er an d norma lly has high an d low pressure ala rm cont acts. See section 4.2.2 which follows. C a u t i o n : Wh e n r e p l a c i n g n i t r o g e n cylinders, do not just order a “n i t r o g e n c y l i n d e r ” fr o m t h e l o c a l w e l d i n g s u p p l i e r . N i t r o g e n fo r t r a n s f o r m e r s s h o u l d m e e t A S TM D-1933 Type III with - 59 °C dew p o i n t a s s p e c i f i e d i n I E EE C-57.12.00-1993, paragraph 6.6.3 [27, 2]. F i g u r e 8 .—P r e s s u r i z e d I n e r t G a s 4.2.2 Gas Pre ssu re Control Transformer. C o m p o n e n t s . A ft e r 1 m o n t h o f s e r v i c e a n d y e a r ly , in s p e c t t h e g a s p r e s s u r e c o n t r o l c o m p o n e n t s . There is norma lly an adjusta ble, thr ee-element pressur e cont rol system for iner t gas, which ma intains a pr essure ra nge of 0.5 to 5 psi in the tr an sform er ta nk. There is also a bleeder valve tha t exhaust s gas to at mosphere when pressur e exceeds relief pres sur e of the valve, norm ally 5 to 8 psi. C a u t i o n : Th e component par t descriptions below ar e for the typical threestage pressure regulating equipment supplying inert gas to th e tr an sform er. Your particular u nit ma y be different, so check your tr an sform er instru ction manual. High Pr essure Gage. The high pressure gage is at ta ched between th e nitr ogen cylinder an d high pressure r egulator th at indicat es cylinder pressure. When the cylinder is full, th e gage will rea d a pproximat ely
Front View
Side View
F i g u r e 9 .—G a s P r e s s u r e C o n t r o l Components.
23
2,400 psi. Norm ally, th e gage will be equipped with a low press ur e alar m th at activat es when t he cylinder is gett ing low (ar ound 500 psi). However, gas will still be supplied, and the regulating equipment will cont inue t o function u ntil th e cylinde r is em pt y. Refer to figur e 9 for th e following descript ions . H i g h P r e s s u r e R e g u l a t o r . The high pressur e regulat or ha s two sta ges. The inpu t of th e first st age is conn ected to th e cylinder , and t he outpu t of th e first sta ge is connected inter na lly to the input of th e second stage. This holds out put pressur e of the second sta ge consta nt . The first sta ge outpu t is adjusta ble by a ha nd-operat ed lever and can deliver a m aximum of wha tever pressur e is in t he cylinder (2,400 psi when full) down to zero. The second st age outpu t is var ied by tu rn ing the a djusting screw, norma lly adjusted t o supply appr oxima tely 10 psi to th e input of the low pressure regulator. L o w P r e s s u r e R e g u l a t o r . The low pressure r egulat or is the th ird stage an d cont rols pressur e an d flow to the gas spa ce of the tr an sform er. The input of this regu lator is conn ected t o th e out put of th e second st age (appr oxima tely 10 psi). This regulator is typically set a t t he factory to supply gas to the t ra nsform er a t a pres sur e of appr oxima tely 0.5 psi and needs no adjust men t. If a different pressur e is required, the r egulator can be adjusted by varying spring tension on th e valve diaphra gm. Pr essure is set at th is low value becau se major pressur e chan ges inside th e tra nsform er come from expan sion an d contr action of oil. The pur pose of this gas feed is to make u p for sm all leaks in t he ta nk gasket s an d elsewhere so th at a ir can not enter . Typically, a spr ing-loaded bleeder for high pressur e relief is built int o the r egulator an d is set a t the factory to relieve pres sur es in excess of 8 psi. The valve will close when press ur e drops below th e setting, preventing further loss of gas. B y p a s s V a l v e A s s e m b l y . The bypass valve assembly opens a bypass line a round th e low pressur e regulator a nd a llows the second st age of the high pressu re regulator to furn ish gas directly to the t ra nsform er. The purpose of this assembly is to allow mu ch fast er filling/pur ging of the gas spa ce during initia l insta llation or if th e tr an sform er t an k ha s to be refilled after being opened for insp ection. C a u t i o n : During norma l operat ion, th e bypass valve must be closed, or pressur e in th e ta nk will be too high. O i l S u m p . The oil sump is locat ed at the bott om of the pr essure r egulating system bet ween th e low pressu re regula tor an d shu toff valve C. The sum p collects oil an d/or moistur e tha t ma y ha ve conden sed in the low pressur e fill line. The dra in plug at th e bottom of the sump should be removed before t he system is put into operat ion a nd a lso rem oved once each year dur ing operat ion t o drain a ny residual oil in t he line. This sump a nd line will be at t he sam e pressure a s th e gas space in th e top of the t ra nsform er. The sump sh ould always be at a safe pressure (less th an 10 psi) so the plug can be r emoved to allow th e line to purge a few seconds a nd blow out th e oil. However, a l w a y s look at th e gas space pressure
24
gage on t he t ra nsform er or th e low pressur e gage in t he n itrogen cabinet, just to be sure, before removing the drain plug. S h u t o f f V a l v e s . The shu toff valves are locat ed nea r t he t op of th e cabinet for t he pur pose of isolating the tr an sform er ta nk for shipping or m ainten an ce. These valves ar e norm ally of double-seat const ru ction a nd sh ould be fully opened a gainst th e stop to prevent gas leaka ge ar oun d the stem . A shu toff valve is also provided for th e purp ose of shu tt ing off th e nitrogen flow to th e tra nsformer t an k. This shu toff valve mu st be closed pr ior t o cha nging cylinders t o keep t he gas in th e transformer tank from bleeding off. S a m p l i n g a n d P u r g e V a l v e. The sampling and purge valve is normally located in th e upper r ight of th e nitr ogen cabinet. This valve is typically equipped with a hose fittin g; th e oth er side is conn ected directly to the tr an sform er gas spa ce by copper t ubing. This valve is opened while purging the gas space durin g a new installation or m ainten an ce refill an d provides a pa th t o exhaust air as t he gas spa ce is filled with nitr ogen. This valve is also opened when a gas sa mple is tak en from t he gas space for a na lysis. When ta king gas sam ples, the line mu st be sufficient ly pur ged so th at th e sam ple will be from gas a bove the t ra nsform er oil an d not just gas in th e line. This valve must be tight ly closed during norma l operat ion t o prevent gas leakage. F r e e B r e a t h i n g C o n s e r v a t o r . This design adds an expansion ta nk (conservator) above the transformer so tha t t he ma in tank may be completely filled with oil. Oil expan sion a nd a ir excha nge with t he at mosphere (breat hing) occurs a way from t he oil in th e tr an sform er. This design redu ces oxygen an d m oisture conta mina tion because only a sma ll port ion of oil is exchan ged between th e main ta nk a nd conservat or. An oil/air inter face still exists in th e conservat or, exposing th e oil to air . Event ua lly, oil in th e conservat or is excha nged with oil in th e main ta nk, an d oxygen and other cont am inates gain access to the insulation.
F i g u r e 1 0 .—F r e e B r e a t h i n g Conservator.
If you h ave tr an sform ers of this design, it is recomm ended th at a bladder or diaph ra gm-type conservat or be inst alled (described below) or ret rofitted t o th e origina l conser vat or. In ad dition, a desiccant -type air dr yer should also be installed. C on s er v a t o r w i t h B l a d d e r o r D i a p h r a g m D e si g n . A conser vator with bladder or diaphr agm is similar to the design above with a n a dded air bla dder (balloon) or flat diaph ra gm in th e conservat or. The bladder or diaph ra gm expan ds and
25
cont ra cts with the oil and isolates it from th e at mospher e. The inside of th e bladder or top of the diaphr agm is open t o atmospheric pressur e through a desiccan t air dryer. As oil expands an d contr acts an d as at mospheric pressur e cha nges, the bladder or diaphra gm “breat hes” air in and out. This keeps air and tr an sform er oil essentially at a tm ospheric pressur e. The oil level gage on t he conser vator typically is Figure 11.—Conservator ma gnetic, like those men tioned earlier, w i t h B l a d d e r. except t he float is positioned near th e cent er of the u n d e r s i d e of the bladder. With a diaphra gm, the level indicat or a rm rides o n t o p of the diaphr agm. Exam ine the air dryer periodically an d cha nge the desiccant when approximately one-third of the material changes color. N o t e : A vacuum will appear in th e tr an sform er if piping between t he a ir dryer an d conser vator is too sma ll, if th e air int ak e to the dr yer is too sma ll, or if th e piping is part ially blocked. The bladder cannot ta ke in air fast en ough when th e oil level is decrea sing due to ra pidly falling temp era tu re. Minium ¾- to 1-inch piping is recomm ended. This problem is especially preva lent with tr an sform ers th at ar e frequ ent ly in an d out of service an d located in geogra phic area s of lar ge temper atu re var iations. This situa tion m ay allow bubbles to form in th e oil an d ma y even activat e gas detector relays such a s th e Buchholz an d/or bladder failure relay. The vacuum m ay also pull in air around gaskets tha t ar e not tight enough or tha t h ave deter iora ted (which m ay also cau se bubbles) [4]. B l a d d e r F a i l u r e ( G a s A c c u m u l a t o r ) R e l a y . The bladder failure r elay (not on diaphragm-type conservators) is mounted on top the conservator for the purpose of detecting air bu bbles in th e oil. Shown at r ight (figure 12) is a modern r elay. Check your tr an sform er VENTS VALVE FLOAT inst ru ction m an ua l for specifics TO DESICCANT ELECTRICAL because designs vary with AIR DRYER CONNECTION ma nufacturer s. No bladder is totally impermea ble, and a little air will migra te into th e oil. In add ition, if a hole forms in th e bladder, allowing air to migrate into th e oil, the r elay will detect BLADDER it. As air rises and enter s the rela y, oil is displaced and t he float drops, activating th e alar m. It is similar t o the top cha mber CONSERVATOR TANK of a Buchholz relay, since it is filled with oil an d cont ain s a float switch. TO TRANSFORMER TANK F i g u r e 1 2 .—B l a d d e r F a i l u r e R e l a y .
26
C a u t i o n : Never open t he vent of the bladder failure r elay unless you h ave vacuum or pressu re equipmen t available. The oil will fall inside the relay and conser vat or and pu ll in air from th e out side. You will ha ve to recomm ission th e rela y by valving off th e conservat or an d press ur izing th e bladder or by placing a vacuu m on th e relay. See your specific tr an sform er instr uction m an ua l for deta ils. C a u t i o n : When th e tra nsform er, relay, an d bladder ar e new, some air or gas is norma lly entr apped in the tr an sform er an d piping and ta kes a while to rise and activat e the relay. Do not assum e the bladder h as failed if the alar m a ctivates within 2 to 3 mont hs a fter it is put into opera tion. If th is occur s, you will ha ve to recomm ision th e relay with press ur e or vacuum . See your specific tr an sform er instru ction ma nua l for deta ils. If no more alar ms occur , the bladder is int act. If ala rm s continu e, look carefully for oil leaks in t he conser vat or and t ra nsform er. An oil leak is usu ally also an air lea k. This ma y be checked by looking at th e nitr ogen an d oxygen in the dissolved gas a na lysis. If th ese gases a re increas ing, th ere is probably a leak; with a sealed conser vator, th ere sh ould be little of th ese gasses in th e oil. Nitr ogen ma y be high if th e tra nsform er was sh ipped new filled with n itrogen. E v e r y 3 t o 5 y e a r s , (if th e conservat or ha s a diaphra gm) remove th e conservat or inspection flange an d look inside with a flash light . If ther e is a leak, oil will be on top of th e diaph ra gm, and it must be replaced. The new diaphr agm ma terial should be nitrile. If the conservat or h as a bladder an d a bladder failur e relay, the relay will alarm if the bladder develops a leak. If the conservat or h as a bladder an d does not ha ve a bladder failur e relay, inspect the bladder by removing the moun ting flan ge and look inside with a flash light . If ther e is oil in th e bott om of th e bladder, a failur e ha s definitely occur red, and the bladder mu st be r eplaced. Follow procedures in t he specific tran sform er instru ction ma nu al for dra ining the conservator and replacement; designs and procedures vary and will not be covered here. A u x i li a r y T a n k S e a l i n g S y s t e m . The auxiliary tan k sealing system incorporates an extra ta nk between the main tra nsformer ta nk and the conservator tank. Inert gas (norm ally nitr ogen) is placed above oil in both th e main an d middle tan ks. Only oil in th e top conservator tank is exposed to air . A desiccant a ir dryer ma y or ma y not be included on th e brea th er. As oil in the main tan k expands and contr acts with temperatu re, gas pressur e varies above the oil in both (figure 13).
F i g u r e 1 3 .—A u x i li a r y S e a l i n g S y s t e m .
27
Chan ges in gas pressur e causes oil to go back and fort h between t he m iddle ta nk an d the conser vat or. Air cont ainin g oxygen and moistu re is not in cont act with oil in the ma in tra nsform er ta nk. Oxygen and moistur e are absorbed by oil in the conser vat or ta nk a nd inter chan ged with oil in the middle one. However, since gas in the middle ta nk inter cha nges with gas in the ma in tan k, small am ount s of oxygen and m oisture carr ied by gas st ill make t heir way into the tr an sform er. With this a rr angemen t, th e conservat or does not have to be locat ed above the ma in ta nk , which reduces the overa ll height . If you ha ve one or more of th ese type tr an sform ers without desiccan t a ir dryers, th ey should be installed.
4.3 Gaskets Gaskets h ave severa l importan t jobs in sealing systems [6]. A gasket m ust creat e a seal an d hold it over a long period of tim e. It mu st be imper vious an d not conta min at e th e insulat ing fluid or gas a bove the fluid. It sh ould be easily rem oved an d replaced. It m ust be elast ic enough to flow int o imper fections on th e sealing surfaces. It m ust withstan d high and low tempera tur es and rem ain resilient enough to hold the seal even with joint movement from expansion, cont ra ction, and vibrat ion. It mu st be resilient en ough to not ta ke a “set” even th ough exposed for a long tim e to pressu re applied with bolt torque and temper at ur e cha nges. It mu st ha ve sufficient strength t o resist crushing un der a pplied load a nd r esist blowout under system pressu re or vacuum . It mu st ma intain its integrity while being ha ndled or installed. If a gasket fails to meet an y of these criteria, a leak will result. Gasket leaks result from improper torque, choosing the wrong type gasket material, or the wrong size gasket. Impr oper sealing sur face prepar at ion or t he gasket ta king a “set” (becoming ha rd a nd losing its resilience and elast icity) will also cause a leak . Usu ally, gasket s ta ke a set as a result of temperat ure extremes and a ge. S e a l i n g ( m a t i n g ) s u r f a c e p r e p a r a t i o n : Clean the meta l surface thoroughly. Remove all moistu re, oil an d grease, ru st, etc. A wire brus h a nd/or solvent m ay be required. Caution: Take extra care that rust and dirt particles ne ver fall into the t r a n s f o r m e r . T h e r e s u l t s c o u l d be c a t a s t r o p h i c , w h e n t h e t r a n s f o r m e r is energized.
After r ust an d scale have been rem oved, meta l surfaces should be coated with Loctite Master gasket No. 518. This mat erial will cur e after you bolt u p the gasket , so addit iona l glue is not necessar y. If th e temper at ur e is 50 °F or more, you can bolt up the gasket immediately. This mat erial comes in a kit (part N o. 22424) with primer, a tu be of ma ter ial, an d instr uctions. If th ese inst ru ctions ar e followed, th e seal will last ma ny years, and t he gasket will be easy to remove later if necessary. If the temper at ur e is under 50 °F, wait about ½ to 1 hour a fter applying the ma terial to sur faces before bolting. If you ar e using cork -nitrile or cork -neopren e, you can also
28
seal gasket su rfaces (including th e edge of the gasket ) with th is same ma terial. Loctite ma kes oth er sealers t ha t can be used to seal gaskets such a s “Hi-tack.” GE glyptol No. 1201B-red can a lso be used to paint gasket an d met al sur faces, but it ta kes more time and you mu st be more cau tious about tem perat ur e. If possible, th is work sh ould be done in tem pera tu res a bove 70 °F to speed paint cur ing. Allow the pain t to completely dry before a pplying glue or t he new gask et. It is not necessar y to rem ove old glyptol or oth er pr imer or old glue if th e sur face is fairly smooth an d uniform. C a u t i o n : Most synt het ic ru bber compoun ds, including nitr ile (Bun a N), conta in some carbon, which ma kes it semi-conductive. Tak e extra car e and n e v e r drop a gasket or pieces of gasket into a tr an sform er ta nk. The results could be cat astr ophic when t he tra nsform er is ener gized.
Choose the corr ect replacement gasket . The ma in influences on gasket m ater ial selection a re design of the gasket joint, ma ximu m a nd m inimum operat ing temper at ur e, type of fluid cont ained, and int erna l pressure of th e tra nsform er. Cork-nitrile should be us ed if th e joint does n o t ha ve grooves or limits. This ma terial perform s better t ha n cork-neoprene becau se it does not tak e a set a s easily an d conform s better to mat ing surfaces. It also perform s better a t higher temper at ur es. Be extr a careful when you store this ma terial becau se it looks like cork -neopren e (described below), and th ey easily ar e mista ken for ea ch oth er. Compr ession is th e sam e as for cork -neopren e, about 45%. Cork -nitrile should recover 80% of its th ickn ess with compr ession of 400 psi in accordan ce with ASTM F36. Ha rdn ess should be 60 to 75 duromet er in accorda nce with ASTM D2240. (See published s pecifications for E -98 by ma nu factu rer Dodge-Regupol Inc., Lan caster , PA.) C a u t i o n : C o r k -n i t r i l e h a s a s h e l f l if e o f o n l y a b o u t 2 y e a r s , s o d o n o t o r d e r a n d s t o c k m o r e t h a n c a n b e u s e d d u r i n g t h i s t im e . C o r k - N e o p r e n e mixtu re (called coropren e) can also be us ed; however, it does not perform as well as cork-nitrile. This mat erial tak es a set wh en it is compr essed and should only be used wh en th ere ar e no expan sion limiting grooves. Using corkneopren e in grooves can resu lt in leaks from expan sion a nd cont ra ction of ma ting sur faces. The ma ter ial is very porous an d should be sealed on both sides and edges with a th in coat of Glyptol No. 1201B red or similar sealer before ins ta lling. Glyptol No. 1201B is a slow drying paint u sed to seal meta l flanges and gaskets, an d th e paint should be allowed to dry totally before insta llation. Once compr essed, this gask et should never be reus ed. These gasket s should be kept a bove 35 °F before inst allat ion to prevent t hem from becoming har d. Gaskets should be cut a nd sealed (painted) indoors a t t empera tu res a bove 70 °F for ea se of handling and to reduce paint curing time. Insta lling neoprene-cork gaskets when tem perat ur es are at or near freezing should be avoided becau se the gasket could be dam aged an d leak . Cork-neopren e gasket s mu st be evenly compr essed about 43 to 45%. For exam ple, if th e gasket is ¼-inch th ick, 0.43 x 0.25 = 0.10. When t he gask et is torqued down, it should be
29
compr essed about 0.10 inch. Or you m ay subt ra ct 0.1 from ¼ inch to calcula te th e th ickn ess of th e gasket a fter it is compr essed. In t his case, ¼ = 0.25 so 0.25 min us 0.10 = 0.15 inch would be the final distan ce between th e ma ting sur faces after th e gasket is compressed . In an emer gency, if compr ession limits ar e required on this gasket, split lock washer s may be used. Bend the washer s unt il th ey ar e flat an d insta ll enough of th em (minium of th ree), evenly spaced, in th e cent er of th e gasket cross section t o preven t excessive compression. The th ickn ess of th e wash ers sh ould be such that the gasket compression is limited to approximately 43%, as explained above. Nitrile “NBR” (Buna N) with 50 to 60 Duro (ha rdn ess) is genera lly the m at erial tha t should be chosen for m ost tra nsform er a pplicat ions. C a u t i o n : Do not confuse th is ma ter ial with But yl Rubber . But yl is not a sat isfactory ma terial for tr an sform er gaskets. The terms But yl an d Buna a re easily confused, an d car e must be taken t o ma ke sure Nitr ile (Buna N) is always used and never But yl.
Replace all cork neoprene gaskets with Nitrile i f t h e j o i n t h a s r e c e s s e s o r e x p a n s i o n l im i t i n g g r o o v e s . Be car eful to protect Nitrile from su nlight; it is not sun light r esista nt a nd will deter iora te, even if only th e edges are exposed. It sh ould not be grea sed when it is used in a nonm ovable (sta tic) seal. When joints h ave to slide dur ing insta llation or a re u sed as a moveable seal (such a s bush ing caps , oil cooler isolation valves, an d ta p cha nger dr ive sha fts), the gask et or O-ring sh ould be lubricated with a t hin coat ing of DOW No. 111 or No. 714 or equivalen t grea se. These ar e very thin an d provide a good seal. Nitrile perform s better t ha n cork-neoprene; when exposed t o higher tem pera tu res, it will perform well up to 65 °C (150 °F). Viton sh ould be used only for gask ets a nd O-rings in t empera tur es higher th an 65 °C or for a pplicat ions r equirin g motion (sha ft seals, etc.). Viton is very tough a nd wea r resist an t; however, it is very expensive ($1,000+ per sh eet) and should not be used un less it is needed for high wear or high tem pera tu re ap plicat ions. Viton should only be used with compr ession limiter grooves an d recesses. Car e should be taken t o store Nitr ile an d Viton separ at ely, or order t hem in different colors; th e ma ter ials look alike an d can be ea sily confused, and a mu ch m ore expensive gasket can be insta lled unn ecessarily. Compr ession a nd fill requiremen ts for Viton a re th e as sa me a s those for n itrile, out lined above and sh own in t able 1. G a s k e t s i z i n g f o r s t a n d a r d g r o o v e d e p t h s . Nitrile is chosen as t he exam ple becau se it is the m ost comm only used ma terial for t ra nsform er gasket ing. As shown in ta ble 1, nitrile compression should be 25 to 50%. Nitr ile sheet s ar e available in 1/16-inch-thick incremen ts.
Gasket t hickness is determ ined by groove depth an d sta ndar d gasket t hickness. Choose the sheet thickness so that one-fourth to one-third of the gasket will protrude above th e groove; th is is the am oun t ava ilable to be compressed . (See ta ble 2.) Gask et sheet s come in sta nda rd th ickn esses in 1/16-inch increm ent s. Choose one th at a llows one-th ird of the ga sket to stick out a bove th e groove if you can, but never choose a
30
Table 1.—Transformer Gasket Application Summary Best Temperature Range
Gasket Material
Percent Compres -sion
Compatible Fluids
UV Resist
Best Applications
Neoprene (use Nitrile except where there is ultraviolet [UV] exposure) or use Viton
-54 to 60 °C (-65 to 140 °F) not good with temp. swings
30 to 33
Askarels and hydrocarbon fluids
Yes
Use only with compression limits or recesses and use only if UV resistance is needed
Cork-Neoprene (Coroprene) this material takes a set easily
0 to 60 °C
40
Mineral oil R-Temp Alpha 1
No
Use only for flat to flat surface gaskets with no grooves or compression limits
40
Mineral oil R-Temp Alpha 1
No
Use only for flat to flat surface gaskets with no grooves or compression limits
25 to 50
Mineral oil
No
O-rings, flat and extruded gaskets; use with compression limiters or recess only
Yes
High temp.; O-rings, flat and extruded gaskets; use with compression limiter groove or recess
(32 to 140 °F)
Cork-Nitrile (best) does not take a set as easily as corkneoprene
-5 to 60 °C
Nitrile (Buna N) use this except in high temp., high wear, or UV
-5 to 65 °C
Viton use for high wear and high temp. applications
-20 to 150 °C
(23 to 140 °F)
(23 to 150 °F)
R-Temp, Alpha 1 Excellent for Hydrocarbon fluids
30 to 33
(-4 to 302 °F)
Silicone, Alpha 1 Mineral oil
Note: Viton O-rings are best for wear resistance and tolerating temperature variations. Nitrile (Buna N) can also be used in low wear applications and temperatures less than 65 °C.
Table 2.—Vertical Groove Compression for Circular Nitrile G askets Standard groove depth (in inches)
Recommended gasket thickness (in inches)
Available to compress (in inches)
Available compression (percent)
3/32
1/8
1/32
25
1/8
3/16
1/16
33
3/16
1/4
1/16
25
1/4
3/8
1/8
33
3/8
1/2
1/8
25
31
thickness th at allows less tha n one-four th or a s mu ch a s one-ha lf to protrude a bove th e groove. Do not t ry to rem ove old primer from t he groove. Horizont al groove fill is deter min ed by how wide the gr oove is. The groove width is equal to th e outer diam eter (OD) minus th e inner diameter (ID) divided by two: OD − ID . Or just measu re the groove width with a n accur at e caliper. 2 The width of th e groove minus t he width of th e gasket is t he r oom left for the gas ket t o expan d while being compr essed. For nitr ile, th e amoun t of horizont al room needed is about 15 t o 25%. Ther efore, you need to cut th e gasket cross section so th at it fills about 75 to 85% of the width of the groove. OD ID
8− 6
− For exam ple, an 8-inch OD groove with a 6-inch ID, is 2 = 1 inch. Ther efore, 2 th e width of th e groove is 1 inch. Because we have to leave 25% expan sion spa ce, the width of th e gasket is 75% of 1 inch, or ¾ inch. So th at th e gasket can expa nd equa lly towar d t he center an d t oward the outside, you should leave one-ha lf the expansion space at th e inn er diameter of th e groove and one-half at t he outer. In th is exam ple, ther e should be a tota l space of 25% of 1 in ch or (¼ in ch) for expansion after the gasket is insert ed, so you should leave �-inch spa ce at the OD and �-inch spa ce at CROSS SECTION OF CIRCULAR GASKET IN GROOVE the ID. See Figu re 14.—Cross Sec tion of Circular Gaske t in Groove . figur e 14.
Al w a y s c u t t h e o u t e r d i a m e t e r f ir s t . In t his example, the outer diameter would be 8 inches m inus ¼ inch, or 7¾ inches. N o t e : Since �-inch space is required all around t he gasket, ¼ inch m ust be subtracted to allow � inch on both sides. The inner dia met er would be 6 inches plu s ¼ inch or 6¼ inches. Note that ¼ inch is su btra cted from the OD but added to the ID.
To check your self, subtra ct t he inner ra dius from th e out er t o make sur e you get t he sam e gasket width calculated a bove. In th is exam ple, 3 -inches (out er r adiu s, ½ of 7¾), min us 3 � inches (inn er r adiu s, ½ of 6¼), is ¾ inch, which is th e corr ect gasket width. R e c t a n g u l a r N i t r il e G a s k e t s larger th an sheet st ock on han d can be fabricated by cutt ing strips and corner s with a ta ble saw or a ut ility knife with r azor blade. Cutt ing is easier if a little tr an sform er oil or WD-40 oil is ap plied. Nitr ile is also available in spools in stan dar d ribbon sizes. The ends may be joined using a cyan oacrylat e
32
adh esive (super glue). Ther e ar e ma ny types of th is glue; only a few of th em work well with nit rile, an d they all have a very limited shelf life. Remem ber to always keep them refrigerat ed to extend shelf life. The one proven to stand u p best to temper at ur e chan ges and compr ession is Lawson Rubber Bonder No. 92081. The Lawson par t nu mber is 90286, and it is ava ilable from La wson P roducts Co. in Reno, Nevada , (702-856-1381). Loctite 404 is comm only ava ilable at NAPA au to par ts st ores a nd work s also but does not sur vive temper at ur e var iat ions as well. Shelf life is crit ical. A new sup ply should a l w a y s be obtained when a gask eting job is star ted; n e v e r use an old bott le tha t h as been on th e shelf since the last job. When bonding th e ends of ribbon t ogeth er, ends sh ould be cut at an an gle (scar fed) at about 15 degrees. The best bond occur s when t he length of th e angle cut is about four t imes th e thickness of th e gask et. With pra ctice, a craftsper son can cut 15-degree scarfs with a u tility knife. A jig can also be ma de from wood to hold the gasket at a 15-degree angle for cutt ing and sa nding. The ends m ay be fur ther fine-sanded or ground on a fine ben ch grinder wheel to ma tch per fectly before applying glue. A jig can be fabricat ed to hold the gasket at 15 degrees while cutting, san ding, or grinding.
Table 3.—Vertical Groove Compression for Rectangular Nitrile Gaskets Standard groove depth (in inches)
Standard ribbon width (in inches)
Recommended gasket thickness (in inches)
Available to compress (in inches)
Available compression (in inches)
3/32
1/4
1/8
1/32
25
1/8
5/16
3/16
1/16
33
3/16
3/8
1/4
1/16
25
1/4
3/4
3/8
1/8
33
3/8
3/4
1/2
1/8
25
N o t e : Maximu m horizont al fill of the gr oove should be 75 to 85% as explained above in th e circular gasket section. However, it is not n ecessa ry to fill th e groove fully to 75% to obta in a good seal. Choose th e width of ribbon t ha t comes close to, but d oes not exceed, 75 to 80%. If one st an dar d r ibbon width fills only 70% of th e groove an d th e next size sta nda rd widt h fills 90%, choose th e size tha t fills 70%. As in t he circular groove explained a bove, place th e gasket so tha t expan sion spa ce is equal on both sides. T h e k e y p o i n t i s t h a t t h e c r o s s s e c t i o n a l a r e a o f t h e g a s k e t r e m a i n s t h e same as the cover is tightened; the thickness dec reases, but the w idth i n c r e a s e s . S e e b e l o w a n d fi g u r e 15 . C a u t i o n : Nitrile (Buna N) is a synthet ic rubber compound a nd, as cover bolts ar e tightened, the gasket is compr essed. Thickness of the gasket is decreased an d th e
33
width is increa sed. If a gask et is too lar ge, ru bber will be pressed in to the void between th e cover an d th e sealing surface. This will prevent a meta l-to-meta l seal, an d a leak will resu lt. It is best if th e cross sectiona l ar ea of the gasket is a little smaller than the groove cross sectional ar ea. As cover bolts are t ighten ed, th e thickness of the gasket decreases but th e width increases so th at cross sectional a rea (thickness times the width) remains th e same. Care must be taken to ensure that the gasket cross sectiona l ar ea is e q u a l t o o r s l i g h t l y s m a l l e r (never larger) tha n th e groove cross sectional ar ea. This will provide space for t he r ubber t o expan d in th e groove so tha t it will not be forced out into the meta l-to-meta l conta ct ar ea. (See figur e 15.) If it is forced out into th e “metal-to-metal” seal area, a leak generally will be the result. When this ha ppens, our first resp onse is t o tight en t he bolts, which bends the cover around th e gasket m at erial in the metal-to-meta l cont act area. The Figu re 15.—Cross Sec tion of Gasket leak m ay st op (or more often not); but th e R e m a i n s C o n s t a n t B e f o r e Ti g h t e n i n g next t ime th e cover is rem oved, gett ing a a n d Af t e r . w x d = g w x g t proper sea l is almost impossible becau se the cover is bent. Take extra car e sizing th e gasket , and t hese pr oblems won’t occur. C a u t i o n : O n s o m e o l d e r b u s h i n g s used on volta ges 15 kV and a bove, it is necessar y to inst all a semicondu ctive gask et. This type bushing (such as GE type L) ha s no groun d connection between the bottom porcelain skirt flange an d th e ground ring. The bottom of th e skirt is norma lly painted with a conductive paint, a nd th en a semicondu ctive gasket is insta lled. This allows stat ic electr ic char ges to bleed off to groun d. The gaskets ar e typically a semiconductive neoprene m at erial. Sometimes, the gasket will have conductive metal st aples near the center to bleed off these cha rges. When replacing this type gasket, always replace with like ma terial. If like gasket m at erial is not a vailable, use cork-neoprene.
Thin m etal conductive shim stock m ay be folded over th e outer per imeter ar ound appr oxima tely one-ha lf th e circumferen ce. These pieces of shim stock sh ould be evenly spaced around th e circum ference and stick far enough in t oward t he cent er so that they will be held when th e bolts ar e tightened. As an example, if the gasket is 8 inches in diam eter , the circum feren ce would be �D or 3.1416 times 8 inches = 25.13 inches in circumferen ce. Fifty percent of 25.13 is about 12½ inches. Cut 12 str ips 1-inch wide and long enough to be clam ped by the flange top an d bott om 34
when t ighten ed. Fold th em over t he outside edge of the gasket leaving a little more th an 1-inch space between, so tha t t he sh im st ock pieces will be more or less evenly spaced ar oun d t he circum ference. N o t e : Fa ilure t o provide a pat h for st at ic electr ic char ges to get t o groun d will resu lt in corona discha rges between t he ground sleeve an d the bush ing flange. The gasket will be ra pidly destr oyed, and a leak will be the r esult. B o l ti n g s e q u e n c e s t o a v o i d s e a l in g p r o b l e m s : If proper bolt tighten ing sequen ces ar e not followed or impr oper torque a pplied to th e bolts, sealing problems will resu lt. The resulting problem is illustr at ed in figure 16. A slight bow in t he flange or lid t op (exaggera ted for illust ra tion) occur s, which applies un even pr essure t o th e gasket. This bow compromises th e seal, and t he gasket will eventu ally leak .
F i g u r e 1 6 .—B o w i n g a t F l a n g e s .
Pr oper bolting sequen ces are illust ra ted for var ious t ype flan ges/covers in figure 17. Bolt numbers show the correct tightening sequences. The n um bers do not h ave t o be followed exactly; however, the dia gona l tighten ing patt erns should be followed. By using proper torque a nd t he illustr at ed sequence patt erns, sealing problems from impr oper tightening an d un even pr essure on th e gasket can be avoided. Use a torque wrench a nd t orque bolts a ccording to th e head stam p on th e bolt. Check ma nu factur ers instru ction book for pan cake gasket t orque values.
4.4 Transforme r Oils 4.4.1 Transforme r Oil Fun ction s. Transformer oils perform at least four functions for t he t ra nsform er. Oil provides insu lation, provides cooling, an d helps extinguish a rcs. Oil also dissolves gases gener at ed by oil degra dat ion, moistur e and gas from cellulose insulation, deterioration, and gases and moisture from wha tever a tm osphere t he oil is exposed to. Close observa tion of dissolved gases in th e oil, an d oth er oil propert ies, provides the most valu able inform at ion a bout transformer health. Looking for trends by comparing information provided in severa l DGAs, and un dersta nding its m eaning, is the m ost importan t tr an sform er dia gnostic tool. 4 .4 .2 D i s s o l v e d G a s An a l y s i s . A ft e r 1 m o n t h o f s e r v i c e a n d o n c e e a c h y e a r , an d more often if a pr oblem is encoun ter ed, do a DGA. This is by far th e most importa nt t ool for deter mining the h ealth of a t ra nsform er.
35
F i g u r e 1 7 .—B o l t T i g h t e n i n g S e q u e n c e s .
C a u t i o n : D G A i s u n r e l i a b l e if t h e t r a n s f o r m e r i s d e -e n e r g i z e d a n d h a s cooled, if the transformer is new , or if it has had less than 1 to 2 w eeks o f continu ous service after oil processin g.
The pu rpose of this section is to provide guidan ce in int erpr eting DGA and t o suggest a ctions based on th e ana lysis. There ar e no “quick an d sur e” easy an swers when dealing with tr an sform ers. Tran sform ers ar e very complex, very expensive, and very important to Reclamation; and each one is different. Decisions m ust be based on experienced judgm ent foun ded on a ll ava ilable data an d consu lta tion with exper ienced people. Along with t horough periodic inspections covered ear lier, the most imp ort an t key to tra nsformer life is periodic DGA an d proper interpr etat ion. Ea ch DGA mu st be compa red to prior DGAs so th at tr ends can be r ecognized an d ra tes of gas generat ion est ablished.
36
A lt h o u g h e x a m p l e s w i l l b e p r e s e n t e d l a t e r , t h e r e i s n o u n i v e r s a l l y a c c e p t e d m e a n s f o r i n t e r p r e t i n g D G A [15]. Tra nsform ers ar e very complex. Aging, chem ical a ctions an d r eactions, electr ic fields, magnet ic fields, ther ma l contraction and expansion, load variations, gravity, and other forces all interact inside the ta nk. Extern ally, thr ough-fau lts, voltage surges, wide am bient temper atu re chan ges, an d other forces such a s th e eart h’s ma gnetic field an d gra vity affect the tr an sform er. Ther e are few if an y “cut an d dried” DGA interpr etat ions; even experts disagree. Consulta tion with oth ers, experience, stu dy, compar ing ear lier DGA’s, keeping a ccur at e records of a tr an sform er’s history, and n oting inform at ion found when a tr an sform er is disassem bled will increas e expertise an d provide life extension to this critical equipm ent . K e e p i n g a c c u r a t e r e c o r d s of each individua l tran sform er is param oun t. If a prior th rough-fau lt, overload, cooling problem, or nea rby light ning st rike h as occur red, th is inform at ion is extrem ely valuable when trying to determ ine what is going on inside the tr an sform er. Baseline tr an sform er test inform at ion should be established when the tr an sform er is new or a s soon a s possible thereafter. This must include DGA, Doble, and other test results, discussed in the testing section, “4.7 Tran sform er Test ing.” 4.4.3 Key Gas Meth od of inter pret ing DGA is set fort h in IEEE [11]. Key gases form ed by degra dat ion of oil and pa per insu lation ar e hydrogen (H 2 ), methane (CH 4 ), eth an e (C 2 H 6 ), ethylene (C 2 H 4 ), a cetylene (C 2 H 2 ), ca rbon mon oxide (CO), an d oxygen (O 2 ). Except for car bon m onoxide an d oxygen, all thes e gases ar e form ed from th e degra dat ion of the oil itself. Car bon m onoxide, carbon dioxide (CO 2 ), and oxygen a re formed from d egra dat ion of cellulose (paper ) insu lat ion. Carbon dioxide, oxygen, nitrogen (N 2 ), and moistur e can a lso be absorbed from the air if th ere is a oil/air int erface, or if th ere is a leak in th e tan k. Some of our tr an sform ers ha ve a pressur ized nitrogen blank et above the oil and, in these cases, nitrogen ma y be near sa tur at ion. (See ta ble 4.) Gas type and am ount s are determ ined by where th e fau lt occur s in the tr ansform er an d the severity and energy of the event. Events r an ge from low energy events su ch a s par tial dischar ge, which pr oduces hydrogen an d tr ace amounts of meth an e and et ha ne, to very high energy sustained arcing, capable of generating all the gases including acetylene, which requires t he m ost energy. 4 .4 .4 T ra n s f o r m e r D i a g n o s i s U s i n g I n d i v i d u a l a n d T o t a l D i s s o l v e d K e y G a s Co n c e n t r a t i o n s . A four -condition, DGA guide t o class ify risk s t o transformers with no previous problems has been developed by the IEEE [11]. The guide us es combina tions of individual gases a nd t ota l combu stible gas concent ra tion. This guide is not un iversally accepted a nd is only one of th e tools used to evalua te tr an sform ers. The four conditions ar e defined below: C o n d i t i o n 1 : Total dissolved combustible gas (TDCG) below this level indicates th e tra nsform er is operat ing satisfactorily. Any individual combust ible gas exceeding specified levels in ta ble 4 should ha ve addit iona l investigat ion.
37
Table 4.—Dissolved Key Gas Concentration Limits in Parts Per Million (ppm) H2
CH4
C2 H2
C2H4
C2H6
CO
CO21
TDCG
Condition 1
100
120
35
50
65
350
2,500
720
Condition 2
101-700
121-400
36-50
51-100
66-100
351-570
2,500-4,000
721-1,920
Condition 3
701-1,800
401-1,000
51-80
101-200
101-150
571-1,400
4,001-10,000
1,921-4,630
Condition 4
>1,800
>1,000
>80
>200
>150
>1,400
>10,000
>4,630
Status
1
CO2 is not included in adding the numbers for TDCG because it is not a combustible gas.
C o n d i t i o n 2 : TDCG within this ra nge indicat es greater tha n n orm al combust ible gas level. Any ind ividua l comb us tible gas exceeding specified levels in ta ble 4 should have add itiona l investigat ion. A fault ma y be presen t. Tak e DGA sam ples at least often enough to calculate the am ount of gas genera tion per da y for each gas. (See ta ble 5 for recommen ded sam pling frequency and actions.) C o n d i t i o n 3 : TDCG within th is ra nge indicates a high level of decomposition of cellulose insu lat ion an d/or oil. Any ind ividua l comb us tible gas exceedin g specified levels in table 4 should have addit iona l investigat ion. A fault or fau lts ar e probably present. Take DGA sam ples at least often en ough to calculat e the am ount of gas genera tion per day for each gas. (See ta ble 5.) C o n d i t i o n 4 : TDCG within th is ra nge indicates excessive decomposition of cellulose insula tion an d/or oil. Cont inu ed opera tion could result in failur e of th e tr an sform er (table 5).
Condition n um bers for dissolved gases given in I EE E C-57-104-1991 (ta ble 4) ar e extremely conservat ive. We have tran sform ers th at h ave operat ed safely with individual gases in Condition 4 with n o problems; however, they ar e sta ble and gases ar e not increa sing, or a re increa sing very slowly. If TDCG and individual gases are increasing significantly (more than 30 ppm/day), the fault is active and th e tr an sform er sh ould be de-energized when Condition 4 levels are reached. A s u d d e n i n c r e a s e i n k e y g a s e s a n d t h e r a t e o f g a s p r o du c t i o n i s m o r e i m p o r t a n t i n e v a l u a t i n g a t r a n s f o r m e r t h a n t h e a m o u n t o f g a s . On e exception is acetylene (C 2 H 2 ). The genera tion of an y amoun t of th is gas a bove a few ppm indicates h igh energy arcing. Tra ce am oun ts (a few ppm) can be generat ed by a very hot ther ma l fau lt (500 °C). A one-time arc cau sed by a n earby light ning str ike or a high-volta ge sur ge can genera te acetylene. If C 2 H 2 is foun d in th e DGA, oil sam ples should be taken weekly to deter mine if additiona l acetylene is being genera ted. If no additiona l acetylene is foun d an d the level is below the IEE E Condition 4, th e tra nsform er ma y cont inue in service. However, if acetylene continu es to increase, th e tr an sform er h as a n a ctive high en ergy
38
Table 5.—Actions Based on Dissolved Combustible Gas
Conditions
TDCG Level or Highest Individual Gas (See Table 4)
TDCG Generation Rates (PPM/Day) <10
Condition 1
720 ppm of TDCG or highest condition based on individual gas from table 4
Condition 2
Condition 3
Condition 4
721-1,920 ppm of TDCG or highest condition based on individual gas from table 4 1,941-2,630 ppm of TDCG or highest condition based on individual gas from table 4 >4,630 ppm of TDCG or highest condition based on individual gas from table 4
Sampling Intervals and Operating Actions for Gas Generation Rates Sampling Interval Annually: 6mo for EHV trans
10-30
Quarterly
>30
Monthly
<10
Quarterly
10-30
Monthly
>30
Monthly
<10
Monthly
10-30
Weekly
>30
Weekly
<10
Weekly
10-30
Daily
>30
Daily
Operating Procedures Continue normal operation.
Exercise caution. Analyze individual gases to find cause. Determine load dependence. Exercise caution. Analyze individual gases to find cause. Determine load dependence.
Exercise extreme caution. Analyze individual gases to find cause. Plan outage. Call manufacturer and other consultants for advice. Exercise extreme caution. Analyze individual gases to find cause. Plan outage. Call manufacturer and other consultants for advice. Consider removal from service. Call manufacturer and other consultants for advice.
NOTES: 1. Either the Highest Condition Based on Individual Gas or Total Dissolved Combustible Gas can determine the condition (1,2,3, or 4) of the transformer [11]. For example, if the TDCG is between 1,941 ppm and 2,630 ppm, this indicates Condition 3. However ,if hydrogen is greater than 1,800 ppm, the transformer is in Condition 4, as shown in table 4.. 2. When the table says “determine load dependence,” this means, if possible, find out if the gas generation rate in ppm/day goes up and down with load. Perhaps the transformer is overloaded. Take oil samples every time the load changes; if load changes are too frequent, this may not be possible. 3. To get TDCG generation rate, divide the change in TDCG by the number of days between samples that the transformer has been loaded. Down-days should not be included. The individual gas generation rate ppm/day is determined by the same method.
intern al ar c an d should be taken out of service. Fu rth er opera tion is extr emely ha zardous and may result in cat ast rophic failure. Operat ing a tra nsform er with an active high energy arc is extr emely hazardous. Table 4 assumes th at no previous DGA tests h ave been made on th e tra nsform er or tha t no r e c e n t hist ory exists. If a previous DGA exists, it sh ould be reviewed to determ ine if th e situa tion is stable (gases a re n ot increa sing significan tly) or unstable (gases are increasing significantly). D e c i d i n g w h e t h e r g a s e s a r e i n c r e a s i n g s i g n i f i c a n t l y d e p e n d s o n y o u r p a r t i c u l a r t r a n s f o rm e r . 39
Compare the current DGA to older DGAs. If the production rate (ppm/day) of any one of the key gases and/or TDCG (ppm) has suddenly gone up, gases are probably increasing significantly. Refer to table 5, which gives suggested a ctions based on total amount of gas in ppm and rate of gas production in ppm/day. Before going to table 5, determine transformer status from table 4; that is, look at the DGA and see if the transformer is in Condition 1, 2, 3, or 4. The condition for a particular transformer is determined by finding the highest level for any individual gas or by using the TDCG [11]. Either the individual gas or the TDCG can give the transformer a higher Condition number, which means it is at greater risk. If the TDCG number shows the transformer in Condition 3 and an individual gas shows the transformer in Condition 4, the transformer is in Condition 4. Always be conservative and assume the worst until proven otherwise. Sampling intervals and recommended actions. When sudden increases occur in dissolved gases, the procedures recommended in table 5 should be followed. Table 5 is paraphrased from table 3 in IEEE C57.104-1991. To make it easier to read, the order has been reversed with Condition 1 (lowest risk transformer) at the top and Condition 4 (highest risk) at the bottom. The table indicates the recommended sampling intervals and actions for various levels of TDCG in ppm. An increasing gas generation rate indicates a problem of increasing severity; therefore, as the generation rate (ppm/day) increases, a shorter sampling interval is recommended. (See table 5.)
Some information has been added to the table from IEEE C57-104-1991; that is, inferred from the text. To see the exact table, refer to the IEEE Standard. If it can be determined what is causing gassing and the risk can be assessed, the sampling interval may be extended. For example, if the core is tested with a megohmmeter and an additional core ground is found, even though table 5 may recommend a monthly sampling interval, an ope rator may choose to lengthen the sampling interval, because the source of the gassing and generation rate is known. A decision should never be made on the basis of just one DGA. It is very easy to contaminate the sample by accidentally exposing it to air. Mislabeling a sample is also a common cause of error. Mislabeling could occur when the sample is taken, or it could be accidentally contaminated or mishandled at the laboratory. Mishandling may allow some gases to escape to the atmosphere and other gases, such as oxygen, nitrogen, and carbon dioxide, to migrate from the atmosphere into the sample. If you notice a transformer problem from the DGA, the first thing to do is take another sample for comparison. In the gas generation chart (figure 18) [13, 16] and discussion below, please note that temperatures at which gases form are o nly approximate. The figure is not drawn to scale and is only for purposes of illustrating temperature relationships, gas types, and quantities. These relationships represent what generally has been proven in controlled laboratory conditions using a mass
40
F i g u r e 1 8 .—C o m b u s t i b l e G a s G e n e r a t i o n V e r s u s T e m p e r a t u r e .
41
spectrometer. This cha rt was used by R.R. Rogers of the Centr al Electr ic Gener at ing Boar d (CEGB) of Englan d to develop the “Rogers Ra tio Method” of an alyzing tr an sform ers (discussed lat er). A vertical ban d at left shows what gases a nd a pproximat e relative qua ntities ar e produced under pa rt ial dischar ge conditions. Note tha t all the gases are given off, but in m uch less quan tity tha n hydrogen. It ta kes only a very low energy event (part ial dischar ge/corona ) to cau se hydr ogen m olecules t o form from t he oil. Gases are formed inside an oil-filled transformer similar to a petroleum refinery still, in t ha t var ious gases begin form ing at specific temper at ur es. Fr om t he Gas Generat ion Ch ar t, we can see relative amounts of gas a s well as a pproximat e temper atu res. Hydrogen and meth an e begin to form in small amounts ar ound 150 °C. Notice from the char t th at beyond ma ximu m point s, metha ne (CH 4 ), etha ne an d ethylene production goes down as tem perat ur e increases. At about 250°C, production of etha ne (C 2 H 6 ) sta rt s. At a bout 350 °C, production of eth ylene (C 2 H 4 ) begins. Acetylen e (C 2 H 2 ) sta rts between 500 °C and 700 °C. In th e past, th e presen ce of only tr ace am oun ts of acetylene (C 2 H 2 ) was consider ed to indicate a t emper at ur e of at least 700 °C ha d occur red; however, recent discoveries ha ve led to the conclusion t ha t a th erm al fault (hot spot) of 500 °C can produce tr ace am ount s (a few ppm). Lar ger am oun ts of acetylene can only be produced above 700 °C by inter na l arcing. Notice tha t between 200 °C and 300 °C, the pr oduction of met ha ne exceeds hydrogen. Sta rt ing about 275 °C and on up, the production of eth an e exceeds met ha ne. At about 450°C, hydr ogen production exceeds all other s un til about 750 °C to 800 °C; then more a cetylene is pr oduced. It sh ould be noted th at sma ll am oun ts of H 2 , CH 4 , and CO ar e produced by norm al aging. Ther ma l decomposition of oil-impr egna ted cellulose produces CO, CO 2 , H 2 , CH 4 , and O 2 . Decom position of cellulose insu lat ion begins a t only about 100 °C or less. Ther efore, operat ion of tr an sform ers at n o more th an 90 °C is imper at ive. Fa ults will produce intern al “hot spots” of far higher tem perat ur es tha n t hese, and th e resultan t gases show up in th e DGA. Table 6 is a char t of “fau lt types,” par ts of which ar e par aph ra sed from the Int ern at iona l Electr otechn ical Comm ission (IEC 60599) [12]. This cha rt is not complet e. It is impossible to cha rt every cau se an d effect du e to th e extrem e complexity of tr an sform ers. DGA mu st be car efully exam ined with th e idea of deter min ing possible faults a nd possible courses of action. These decisions ar e based on judgm ent a nd experience and ar e seldom “cut an d dried.” Most professiona l associations a gree tha t t here a re t wo basic fau lt types, therm al an d electr ical. The first th ree on th e cha rt a re electr ical discha rges, and the last t hr ee are th ermal faults. Et ha ne an d ethylene are sometimes called “hot meta l gases.” When th ese gases ar e being generat ed an d acetylene is not, the problem found inside the tr an sform er norma lly involves hot meta l. This may include bad conta cts on the ta p changer or a bad conn ection somewhere in t he circuit, such a s a ma in tr an sform er lead. Str ay flux impinging on th e tan k (such as in Westinghouse 7M series tr an sform ers) can cau se these “hot meta l gases.” A shield has been known
42
to become loose an d fall and become un grounded. Sta tic can th en build up and dischar ge to a grounded sur face an d produce “hot met al” gases. An un inten tional core ground with circulating cur rent s can also produce these gases. There ar e ma ny other exam ples. Notice that both type faults (thermal and electrical) may be occurring at once, and one ma y cau se the other. The associations do not m ention m agnetic fau lts; however, magnetic fau lts (such as st ra y ma gnetic flux impinging the st eel tank or other ma gnetic str uctures) also cau se hot spots. A t m o s p h e r i c g a s s e s (N 2 , C O 2 , and O 2 ) can be very valua ble in a DGA in revea ling a possible leak. However, as men tioned elsewher e, th ere ar e oth er reasons th ese gases a re found in DGA. Nitrogen ma y have come from sh ipping the t ran sformer with N 2 inside or from a nitr ogen blan ket . CO 2 a n d O 2 ar e form ed by degrada tion of cellulose. Be very careful; look a t sever al DGAs, and see if at mospher ic gases an d possibly moistur e levels ar e increa sing. Also look at th e tr an sform er carefully if you can find a n oil leak. Moistu re an d atm ospheric gases will leak inside when t he tr an sform er is off an d am bient t empera tu re drops. (See section 4.3 on moistur e) D i s s o lv e d g a s s o f t w a r e . Severa l compa nies offer DGA comput er softwa re t ha t diagnose tra nsform er problems. These diagnoses mu st be used with en gineering judgment a nd should never be taken at face value. The softwar e is consta ntly chan ging. The Technical Service Cent er u ses “ Tra nsform er Oil Ana lyst” (TOA) by Delta x Research. This softwar e uses a composite of severa l curr ent DGA met hods. Dissolved gas an alysis help is available from th e TSC at D-8440 an d D8450. Both groups have th e above softwar e and experience in diagnosing tr an sform er problems.
One set of ru les tha t TOA uses t o genera te a lar ms is based loosely on IEC 60599 (table 6). These ru les are also very useful in daily dissolved gas an alysis. They ar e based on L1 limits of IEC 60599 except for a cetylene. IEC 60599 gives a ra nge for L1 limit s instea d of a specific value. TOA uses t he aver age in this ra nge an d th en gives the u ser a “hea ds up ” if a gener at ion r at e exceeds 10% of L1 limit s per mont h. Acetylene is the exception; IEEE sets a n L1 limit of 35 ppm (too high), an d IEC sets a cetylene ra nge at 3 t o 50. TOA picks th e lowest nu mber (3 ppm) and sets th e generation ra te alarm value at 3 ppm per month. N o t e s : If one or m ore gas genera tion ra tes a re equ al to or exceed G1 limits (10% of L1 limit s per m onth ), you sh ould begin to pay more a tt ent ion t o th is tr an sform er. Reduce the DGA sam ple inter val, reduce loading, plan for futu re outa ge, conta ct t he ma nu factur er etc.
If one or m ore combu stible gas genera tion ra tes a re equ al to or exceed G2 limits (50% of L1 limit s per m onth ), th is tra nsform er sh ould be considered in critical condit ion. You may wan t to reduce sam ple int ervals to month ly or weekly, plan an outa ge, plan to rebuild or r eplace the t ra nsform er, etc. If an active ar c is
43
Table 6.—TOA L1 Limits and Generation Rate Per Month Alarm Limits GAS
L1 Limits
G1 Limits (ppm per month)
G2 Limits (ppm per month)
H2
100
10
50
CH4
75
8
38
C2H2
3
3
3
C2H4
75
8
38
C2H6
75
8
38
CO
700
70
350
CO2
7,000
700
3,500
present (C 2 H 2 genera tion), or if oth er h eat gases a re h igh (above Condit ion 4 limits in ta ble 4), an d G2 limits ar e exceeded, th e tra nsform er should be rem oved from service. Table 7 is ta ken from I EC 60599 of key gases, possible faults, a nd possible findin gs. This cha rt is not all inclusive and sh ould be used with oth er inform at ion. Additional possible fau lts a re listed on following and preceding pages. Tran sform ers a re so complex tha t it is impossible to put a ll sympt oms an d causes into a cha rt. Severa l additiona l tran sform er problems ar e listed below; ther e are man y others. 1. Gases ar e genera ted by norm al operat ion a nd aging, mostly H 2 an d CO with some CH 4 . 2. Operat ing tra nsform ers a t sust ained overload will generat e combust ible gases. 3. Pr oblems with cooling systems, discussed in a n ear lier section, can cau se overheating. 4. A blocked oil duct inside th e tr an sform er can cau se local overhea ting, generat ing gases. 5. An oil directing baffle loose inside th e tr an sform er cau ses mis-direction of cooling oil. 6. Oil circulat ing pum p problems (bear ing wear, impeller loose or worn) can cause tr an sform er cooling problems. 7. Oil level is too low; this will not be obvious if th e level ind icator is in opera tive. 8. Sludge in the tr an sform er and cooling system . (See “3. Tra nsform er Cooling Methods.”)
44
Table 7.—Fault Types Key Gases H2, possible trace of CH 4 and
Possible Faults
Possible Findings
Partial discharges (corona)
Weakened insulation from aging and electrical stress.
H2, CH4, (some CO if discharges involve paper insulation). Possible trace amounts of C 2 H6.
Low energy discharges (sparking).
Pinhole punctures in paper insulation with carbon and carbon tracking. Possible carbon particles in oil. Possible loose shield, poor grounding of metal objects
H2, CH4, C2 H6, C2H4, and the key gas for arcing C 2 H2 will be present perhaps in large amounts. If C 2 H2 is being generated, arcing is still going on. CO will be present if paper is being heated.
High energy discharges
H2, CO.
Thermal fault less than 300 °C in an area close to paper insulation (paper is being heated).
Discoloration of paper insulation. Overloading and or cooling problem. Bad connection in leads or tap changer. Stray current path and/or stray magnetic flux.
H2, CO, CH4, C2H6, C2 H4.
Thermal fault between 300 °C and 700 °C
Paper insulation destroyed. Oil heavily carbonized.
All the above gases and acetylene in large amounts.
High energy electrical arcing
Same as above with metal discoloration. Arcing may have caused a thermal fault.
C2H6. Possible CO.
(May be static discharges)
(arcing)
700 °C and above.
Metal fusion, (poor contacts in tap changer or lead connections). Weakened insulation, from aging and electrical stress. Carbonized oil. Paper destruction if it is in the arc path or overheated.
9. Circulating stra y cur rent s may occur in th e core, structur e, an d/or ta nk. 10. An u ninten tiona l core ground ma y cau se heat ing by providing a pa th for st ra y currents. 11. A hot-spot can be cau sed by a bad conn ection in t he leads or by a poor cont act in the ta p chan ger. 12. A hot-spot m ay also be caused by discha rges of sta tic electr ical char ges tha t build up on shields or core a nd st ru ctu res which ar e not pr operly grounded. 13. Hot-spots ma y be cau sed by electr ical ar cing between windings and groun d, between win dings of differen t potent ial, or in a rea s of different potent ial on t he sam e winding, due to deteriora ted or dam aged insulation. 14. Windings and insulat ion can be dama ged by fau lts downstr eam (thr ough fau lts), cau sing lar ge cur rent surges thr ough th e windings. Through fau lts cause extreme m agnetic and physical forces tha t can distort an d loosen windings and
45
wedges. The result m ay be arcing in the tr an sform er, beginning at th e time of the fau lt, or t he insulat ion m ay be weakened a nd a rcing develop later. 15. Insulat ion can a lso be dama ged by a voltage surge such as a near by light ning str ike or switching sur ge or closing out of step, which ma y result in im media te ar cing or ar cing tha t develops later . 16. Insulat ion ma y be deteriora ted from age and simply worn out. Clear an ces an d dielectr ic str ength ar e redu ced, allowing part ial discha rges an d a rcing to develop. This can also reduce physical stren gth a llowing wedging an d windings to move extensively during a through-fault, causing total mechanical and electrical failure. 17. High noise level (hum d ue t o loose windings) can genera te gas du e to heat from friction. Compa re th e noise to sister t ra nsformer s, if possible. Sound level meter s ar e available at the TSC for diagnostic compa rison a nd t o esta blish baseline n oise levels for futu re compar ison. T e m p e r a t u r e . Gas production r ates increase exponent ially with temper at ur e, an d directly with volume of oil and paper insulation a t h igh en ough t empera tur e to produce gases [11]. Temperat ur e decreases a s distan ce from th e fault increases. Temperat ur e at th e fau lt center is highest, an d oil an d paper her e will produce the most gas. As distan ce increases from the fau lt (hot spot), tem pera tu re goes down an d th e ra te of gas genera tion also goes down. Because of th e volume effect, a lar ge heat ed volume of oil an d paper will produce the sa me am ount of gas as a smaller volume at a h igher tem perat ur e [11]. We can not tell th e difference by looking at t he DGA. This is one r eason th at inter pret ing DGAs is not a n exa ct science. G a s M i x i n g . Concent ra tion of gases in close proximity t o an active fau lt will be higher t ha n in th e DGA oil sample. As distan ce increases from a fau lt, gas concent ra tions decrease. Equ al m ixing of dissolved gases in th e tota l volume of oil depend s on tim e an d oil circulat ion. If th ere ar e no pum ps to force oil th rough ra diat ors, complet e mixing of gases in t he tota l oil volum e ta kes longer. With pumping and normal loading, complete mixing equilibrium should be reached within 24 hours an d will ha ve litt le effect on DGA if an oil sam ple is ta ken 24 hours or m ore a fter a problem begins. G a s S o l u b i l i t y . Solubilities of gases in oil vary with tem perat ur e and pr essure [13]. Solubility of all tr an sform er gases va ry pr oportionally up an d down with p r e s s u r e . Var iat ion of solubilities with t e m p e r a t u r e is mu ch more complex. Solubilities of hydr ogen, nit rogen, carbon m onoxide, an d oxygen go up an d down proportionally with tem pera tu re. Solubilities of carbon dioxide, acetylene, ethylene, an d etha ne ar e reversed an d vary inversely with tempera tu re cha nges. As temperatur e r i s e s , solubilities of th ese gases go down; an d as t emper at ur e falls , their solubilities increase. Metha ne solubility rem ains a lmost consta nt with temper at ur e cha nges. Table 7 is accur at e only at s t a n d a r d t e m p e r a t u re a n d pres sure (STP), (25 °C/77 °F) an d (14.7 psi/29.93 inch es of me rcury, w hich i s s t a n d a r d b a r o m e t r i c p r e s s u r e a t s e a l e v e l ). Table 8 shows only relative differences in h ow gases dissolve in tr an sform er oil.
46
Fr om th e solubility ta ble 8 below, compa rin g hydrogen with a solubility of 7% an d acetylene with s olubility of 400%, you can see th at tr an sform er oil ha s a m uch grea ter capacity for dissolving acetylene. However, 7% hydr ogen by volum e repr esent s 70,000 ppm , an d 400% acetylene repr esent s 4,000,000 ppm. You will probably never see a DGA with nu mbers t his high. Nitrogen can a pproach ma ximum level if th ere is a pressur ized nitr ogen blank et above th e oil. Table 8 shows the m a x i m u m am ount of each gas t ha t t he oil is capa ble of dissolving at sta nda rd tem perat ur e and pressur e. At th ese levels, the oil is said to be saturated.
Table 8.—Dissolved Gas Solubility in Transformer Oil Accurate Only at STP, 25 °C (77 °F) and 14.7 psi (29.93 inches of mercury)
Formula
Solubility in Transformer Oil (% by Volume)
Equivalent (ppm by Volume)
Hydrogen1
H2
7.0
70,000
Partial discharge, corona, electrolysis of H 2O
Nitrogen
N2
8.6
86,000
Inert gas blanket, atmosphere
Carbon Monoxide1
CO
9.0
90,000
Overheated cellulose, air pollution
Oxygen
O2
16.0
160,000
Atmosphere
Methane1
CH 4
30.0
300,000
Overheated oil
Carbon Dioxide
CO2
120.0
1,200,00
Overheated cellulose, atmosphere
Ethane1
C 2H 6
280.0
2,800,00
Overheated oil
Ethylene1
C 2H 4
280.0
2,800,000
Very overheated oil
Acetylene1
C 2H 2
400.0
4,000,000
Arcing in oil
Dissolved Gas
1
Primary Causes/Sources
Denotes combustible gas. Overheating can be caused both by high temperatures and by unusual or abnormal electrical stress.
If you have conservator-type transformers and nitrogen, oxygen, and CO 2 a r e increas ing, ther e is a good possibility th e ta nk h as a leak, or the oil ma y have been poorly processed. Check th e diaph ra gm or bladder for leak s (section 4.2), and check for oily residu e around th e pressur e relief device and oth er gasket ed openin gs. Ther e should be fairly low n itrogen a nd especially low oxygen in a conser vator-type tr an sform er. However, if the tr an sform er was shipped new with pressur ized nitr ogen inside and ha s not been degassed properly, there m ay be high nitrogen cont ent in t he DGA, b u t t h e n i t r o g e n l e v e l s h o u l d n o t b e i n c r e a s i n g after the tr an sform er ha s been in service for a few year s. When oil is being insta lled in a new tr an sform er, a vacuum is placed on t he ta nk wh ich pulls out nitrogen a nd pu lls in th e oil. Oil is free to absorb nitr ogen a t th e oil/gas inter face, and some nitr ogen ma y be trapped in the windings, paper insulat ion, and str uctur e. In th is case, nitrogen ma y be fair ly high in t he DGAs. However, oxygen should be very low, and nitr ogen should not be increasing. It is import an t to take a n oil sample early in th e 47
tr an sform er’s ser vice life to esta blish a baseline DGA; t h e n t a k e s a m p l e s a t l e a s t a n n u a l l y . The nit rogen an d oxygen can be compa red with ear lier DGAs; an d if th ey increas e, it is a good indication of a leak . If th e tr an sform er oil has ever been de gassed, nitr ogen an d oxygen should be low in th e DGA. It is extr emely import an t to keep a ccur at e records over a tr an sform er’s life; when a pr oblem occurs, r ecorded information helps greatly in troubleshooting. 4.4.5 Roge rs Ratio Meth od of DGA. Rogers Ra tio Meth od of DGA [18] is an add itiona l tool th at ma y be used t o look at dissolved gases in tr an sform er oil. Rogers Rat io Meth od compa res qu an tities of different key gases by dividing one int o th e other . This gives a rat io of th e amoun t of one key gas to an other . By looking at t he Gas Genera tion Cha rt (figur e 18), you can see th at , at certain t empera tur es, one gas will be genera ted more than an oth er gas. Rogers used these relationships and determ ined tha t if a certa in ra tio existed, then a specific temper at ur e had been reached. By compa ring a large number of tr an sform ers with similar gas r at ios and dat a found wh en th e tra nsform ers were examined, Rogers could then say tha t certa in fau lts were present. Like the Key Gas Analysis above, this method is not a “sur e thin g” an d is only an a dditional tool to use in a na lyzing tr an sform er pr oblems. Rogers Ra tio Meth od, using th ree-key gas r at ios, is based on ear lier work by Doern ebur g, who used five-key gas ra tios. Rat io met hods ar e only valid if a significan t a moun t of th e gases used in th e ra tio is present . A good ru le is: N e v e r make a decision based only on a ratio if either of the two gases use d in a r a t io i s l e s s t h a n 1 0 t i m e s t h e a m o u n t t h e g a s c h r o m a t o g ra p h c a n d e t e c t (12). (Ten t imes th e individua l gas det ection limits a re sh own in t able 9 and below.) This rule makes sur e tha t instr um ent inaccur acies have little effect on th e rat ios. If either of th e gases a re lower th an 10 times t he det ection limit, you m ost likely do not ha ve th e part icular pr oblem th at t his rat io deals with a nyway. If the gases are not at least 10 times t hese limits, th is does not mea n you can not use t he Rogers Rat ios; it means th at t he results are not as certa in as if the gases were at least a t th ese levels. This is another r eminder th at DGAs ar e not a n exact science and t here is no “one best easy way” to ana lyze tr an sform er pr oblems. Appr oxima te detection limits ar e as follows, depending on th e lab an d equipment: D i s s o l v e d G a s A n a l y s i s D et e c t i o n L i m i t s .
Hydr ogen (H 2 ) about 5 ppm Metha ne (CH 4 ) about 1 ppm Acet ylene (C2 H 2 ) about 1 to 2 ppm Et hylene (C 2 H 4 ) about 1 pp m Etha ne (C 2 H 6 ) about 1 ppm Carbon monoxide (CO) and carbon cioxide (CO 2 ) about 25 ppm Oxygen (O 2 ) an d nitr ogen (N 2 ) about 50 ppm When a fau lt occurs inside a tr an sform er, ther e is no problem with minium gas am ount s at which the ra tio ar e valid. There will be more tha n enough gas present.
48
If a tr an sform er h as been operating n orm ally for some time a nd a DGA shows a sudden increase in t he a mount of gas, th e first t hing to do is take a second sa mple to verify there is a pr oblem. Oil sam ples are easily conta mina ted dur ing sampling or a t th e lab. If the next DGA shows gases to be more in line with pr ior DGAs, th e ear lier oil sam ple was cont am inat ed, and th ere is no fur ther cause for concern. If the second sa mple also shows increa ses in gases, the pr oblem is real. To apply Rat io Methods, it helps to subtract gases th at were present pr ior t o sudden gas increases. This takes out gases tha t ha ve been generated u p to this point due to norm al aging an d from prior problems. This is especially tr ue for r at ios using H 2 an d th e cellulose insulation gases CO and CO 2 [12]. These are genera ted by norm al aging. R o g e r s R a t i o M et h o d U s e s t h e Fo l l o w i n g T h r e e R a t i o s .
C 2 H 2 /C 2 H 4 ,
CH 4 /H 2, C 2 H 4 /C 2 H 6
These rat ios an d the r esultan t fault indicat ions a re based on large num bers of DGAs an d tra nsform er failur es and wha t was discovered after t he failures. Ther e ar e oth er r at io met hods, but only th e Rogers Rat io Meth od will be discussed since it is th e one most comm only used. The met hod descript ion is para phr ased from Rogers’ original paper [18] and from IEC 60599 [12]. C a u t i o n : Rogers Ratio Method is for f a u l t a n a l y z i n g , n o t f o r fa u l t d e t e c t i o n . You m u s t h a v e a l r e a d y d e c i d e d tha t you h ave a pr oblem from the t ota l amount of gas (using IEE E limits) or increa sed gas gener at ion r at es. Rogers Rat ios will only give you a n ind icat ion of wha t t he pr oblem is; it c a n n o t tell you wh eth er or n ot you ha ve a pr oblem. If you alr eady suspect a problem based on tota l combu stible gas levels or increased r at e-of-gener at ion, th en you will norma lly alr eady ha ve enough gas for t his met hod to work . A good system t o deter mine wh eth er you ha ve a problem is to use t able 5 in the Key Gas Meth od. If two or m ore of th e key gases ar e in condition two and t he gas generat ion is a t least 10% per month of the L1 limit, you h ave a p roblem. Also, for t he dia gnosis to be valid, gases used in ra tios should be at least 10 times the detection limits given ear lier. The more gas you ha ve, th e more likely the Rogers Ra tio Method will give a va lid diagn osis. The r everse is also tr ue; the less gas you h ave, th e less likely th e diagnosis will be valid. If a gas used in th e denomina tor of an y rat io is zero, or is shown in th e DGA as n ot det ected (ND), use th e detection limit of tha t pa rticular gas as t he denominator. This gives a rea sonable rat io to use in diagnostic ta ble 9. Zero codes mea n th at you do not ha ve a problem in th is area.
49
Table 9.—Rogers Ratios for Key Gases Code range of ratios
C 2H2 C2H4
CH4 H2
C2H4 C2H6
<0.1 0.1-1 1-3 >3
0 1 1 2
1 0 2 2
0 0 1 2
Detection limits and 10 x detection limits are shown below: C2H2 1 ppm 10 ppm C2H4 1 ppm 10 ppm CH4 1 ppm 10 ppm H2 5 ppm 50 ppm C2H6 1 ppm 10 ppm
Case
Fault Type
Problems Found
0
No fault
0
0
0
1
Low energy partial discharge
1
1
0
Electric discharges in bubbles, caused by insulation voids or super gas saturation in oil or cavitation (from pumps) or high moisture in oil (water vapor bubbles).
2
High energy partial discharge
1
1
0
Same as above but leading to tracking or perforation of solid cellulose insulation by sparking, or arcing; this generally produces CO and CO2.
3
Low energy discharges, sparking, arcing
1-2
0
1-2
4
High energy discharges, arcing
1
0
2
Discharges (arcing ) with power follow through; arcing breakdown of oil between windings or coils, or between coils and ground, or load tap changer arcing across the contacts during switching with the oil leaking into the main tank.
5
Thermal fault less than 150 °C (see note 2)
0
0
1
Insulated conductor overheating; this generally produces CO and CO 2 because this type of fault generally involves cellulose insulation.
6
Thermal fault temp. range 150-300 °C (see note 3)
0
2
0
7
Thermal fault temp. range 300-700 °C
0
2
1
Spot overheating in the core due to flux concentrations. Items below are in order of increasing temperatures of hot spots. Small hot spots in core. Shorted laminations in core. Overheating of copper conductor from eddy currents. Bad connection on winding to incoming lead, or bad contacts on load or no-load tap changer. Circulating currents in core; this could be an extra core ground, (circulating currents in the tank and core); this could also mean stray flux in the tank.
8
Thermal fault temp. range over 700 °C (see note 4)
0
2
2
Normal aging
Continuous sparking in oil between bad connections of different potential or to floating potential (poorly grounded shield etc); breakdown of oil dielectric between solid insulation materials.
These problems may involve cellulose insulation which will produce CO and CO2.
Notes: 1. There will be a tendency for ratio C2H2 /C2H4 to rise from 0.1 to above 3 and the ratio C 2H4 /C2H6 to rise from 1-3 to above 3 as the spark increases in intensity. The code at the beginning stage will then be 1 0 1. 2. These gases come mainly from the decomposition of the cellulose which explains the zeros in this code. 3. This fault condition is normally indicated by increasing gas concentrations. CH4 /H2 is normally about 1, the actual value above or below 1, is dependent on many factors such as the oil preservation system (conservator, N 2 blanket, etc.), the oil temperature, and oil quality. 4. Increasing values of C2H2 (more than trace amounts), generally indicates a hot spot higher than 700 °C. This generally indicates arcing in the transformer. If acetylene is increasing and especially if the generation rate is increasing, the transformer should be deenergized, further operation is extremely hazardous. General Remarks:
1. Values quoted for ratios should be regarded as typical (not absolute). This means that the ratio numbers are not “carved in stone”; there may be transformers with the same problems whose ratio numbers fall outside the ratios shown at the top of the table. 2. Combinations of ratios not included in the above codes may occur in the field. If this occurs, the Rogers Ratio Method will not work for analyzing these cases. 3. Transformers with on-load tap changers may indicate faults of code type 2 0 2 or 1 0 2 depending on the amount of oil interchange between the tap changer tank and the main tank.
50
Example 1 E x a m p l e of a Reclamation transformer DGA: Rogers Ratio Analysis
Hydr ogen (H 2 ) Metha ne (CH 4 ) Etha ne (C 2 H 6 ) Ethylene (C 2 H 4 ) Acetylen e (C 2 H 2 ) Carbon Monoxide (CO) Carbon Dioxide (CO 2 ) Nitr ogen (N 2 ) Oxygen (O 2 ) TDCG
9 ppm 60 53 36 8 3 7 36 1 86,027 1,177 500
Code 0
C2H2 /C2H4 = 3/368 = 0.00815 CH4 /H2 = 60/9 = 6.7
2
C2H4 /C2H6 = 368/53 = 6.9
2
This code combination is Case 8 in table 4, which indicates this transformer has a thermal fault hotter than 700 °C.
Et hylene a nd eth an e ar e sometimes called “hot meta l gases.” Notice th is fau lt does not involve paper insula tion, becau se CO is very low. H 2 a n d C 2 H 2 ar e both less than 10 times the detection limit. This means the diagnosis does not have a 100% confidence level of being corr ect. However, due t o the h igh eth ylene, the fault is pr obably a ba d conn ection wh ere a n incomin g lead is bolted t o a windin g lead, or perha ps bad ta p chan ger cont acts, or additiona l core groun d (lar ge circulat ing cur rent s in th e tan k an d core). See the t wo bott om pr oblems on ta ble 10 lat er in th is cha pter. This exam ple was chosen to show a tr an sform er th at was n ot a “clear cut” diagnosis. Engineering judgment is always required. A small qua nt ity of acetylene is presen t, just a bove the det ection limit of 1 ppm. This is not high energy ar cing due to th e sma ll amount; it ha s m ore likely been produced by a one-time nea rby light ning st rike or a volta ge surge. Example 2 Latest DGA Hydrogen (H2)
Prior DGA No. 2
Prior DGA No. 1
26 ppm
27
17
Methane (CH4)
170
164
157
Ethane (C2H6)
278
278
156
Ethylene (C2H4)
25
4
17
Acetylene (C2H2)
2
0
0
92
90
96
3,125
2,331
2,476
67,175
72,237
62,641
608
1,984
440
Carbon Monoxide (CO) Carbon Dioxide (CO2) Nitrogen (N2) Oxygen (O2)
51
Rogers Ratio Analysis Based on Latest DGA: Codes C2H2 /C2H4
= 2/25
= 0.080
0
CH4/ H2
= 170/26
= 6.54
2
C2H4 /C2H6
= 25/278
= 0.09
0
Notice tha t m etha ne is increasing slowly, but eth an e ha d a large increase between sam ples 1 and 2 but did not increase between sam ples 2 and 3. Note that two key gases (CH 2 a n d C 2 H 6 ) ar e above IEE E Condition 1 in ta ble 5, so the Rogers Ra tio Meth od is valid. By referrin g to table 9, th is combinat ion of codes is Cas e 6, which indicat es the tra nsform er ha s a ther ma l fau lt in th e tempera tur e ran ge of 150 °C to 300 °C. Life history of th e tra nsformer m ust be exam ined car efully. It is, aga in, very importa nt to keep accur at e records of every tr an sform er. This inform at ion is invaluable when it becomes n ecessary t o do an evaluation. The tra nsform er in th is exam ple is one of th ree sister tr an sform ers th at h ave had increased cooling insta lled an d a re r un ning higher loads due t o a genera tor upgra de several years a go. Tran sform er sound level (hu m) is ma rkedly higher th an for th e two sister tra nsform ers. The unit break er experienced a fau lt some years a go, which placed high m echa nical str esses on t he tr an sform er. This genera lly mea ns loose windings, which can gener at e gas du e to friction (called a th erma l fau lt) by Rogers Ratios. Compa rison with sister units r eveals almost tr iple the eth an e as t he other t wo, and it is above the IE EE Condition 4.Gases are increasing slowly; there ha s been no sudden ra te increase in combust ible gas production. Notice the large increa se in O 2 a n d N 2 between th e first an d second DGA an d the lar ge decrease between th e second an d third. This probably mean s th at th e oil sam ple was exposed to air (at mosphere) an d th at th ese two gases are inaccur at e in the m iddle sample. C a r b o n D i o x i d e C a r b o n M on o x id e R a t i o . This ra tio is not included in t he Rogers Ra tio Meth od of an alysis. However, it is useful to deter min e if a fau lt is affecting th e cellulose insulat ion. This ra tio is included in tr an sform er oil an alyzing softwar e programs such as Delta X Research Tra nsform er Oil Ana lyst. This an alysis is ava ilable from t he TSC at D-8440 and D-8450 in Denver.
Form at ion of CO 2 an d CO from th e degradat ion of oil impr egnated pa per increases rapidly with temperature. CO 2 /CO ra tios less tha n t hr ee ar e generally considered an indication of probable paper involvement in an electr ical fault (ar cing or spa rk ing), along with some car bonizat ion of pap er. Norm al CO 2 /CO ra tios a re typically around seven. Ratios above 10 generally indicate a thermal fault with the involvement of cellulose. T h i s i s o n l y t r u e i f t h e C O2 c a m e f r o m w i t h i n the transformer (no leaks), and these ratios are only mean ingful if there i s a s i g n i f i c a n t a m o u n t o f b o t h g a s e s . Cau tion must be employed becau se oil 52
degradat ion a lso produces th ese gases, and CO 2 can a lso be dissolved in t he oil from a tm ospheric leaks. The oil sample can also pick up CO 2 a n d O 2 if it is exposed to air du ring sam pling or ha ndling at t he lab. If a fault is suspected, look carefully to see if CO is increasin g. If CO is increa sing ar ound 70 ppm or more per mont h (gener at ion limit from IE C 60599), ther e is probably a fau lt. It is a good idea to subtr act the a mount of CO an d CO 2 shown before t he increase in CO and CO 2 began , so tha t only gases cau sed by the pr esent fault a re u sed in the r at io. This will eliminat e CO and CO 2 generat ed by norm al aging and other sour ces. When excessive cellulose degra dat ion is sus pected (CO 2 /CO ra tios less th an 3, or greater th an 10), it m ay be advisable to ask for a fura n a na lysis with t he next DGA. This will give an indication of useful life left in th e paper insu lation [12]. You cann ot de-energize a tr an sform er based on fura n an alysis alone. All th is test does is g i v e a n i n d i c a t io n of the h ealth of the pa per; it is not a sure t hing. But fur an an alysis is recomm ended by ma ny experts t o give an indicat ion of remaining life when t he CO 2 /CO ratio is less tha n 3 or great er th an 10. Some oil labora tories do this test on a routine basis, and some charge extra for it. Table 10 is ada pted from I EC 60599 Appendix A.1.1 [12]. Some of th e wording ha s been changed to reflect American lan guage usage rath er th an E ur opean.
4.5 Moisture Pro blems
Moistu re, especially in the pr esence of oxygen, is extrem ely hazar dous to tra nsform er insula tion. Ea ch DGA an d Doble test resu lt should be exam ined car efully to see if water is increasing and to determine the moisture by dry weight (M/DW) or percent satu ra tion th at is in th e paper insulat ion. When 2% M/DW is reached, plans should be ma de for a dry out. Never a llow th e M/DW to go above 2.5% in t he pa per or 30% oil satu ra tion without drying out t he tr an sform er. Ea ch time th e moisture is doubled in a tr an sform er, th e life of th e insulat ion is cut by one-ha lf. Keep in m ind th at t he life of the t ra nsform er is th e life of the paper , and t he pur pose of the paper is to keep out moistur e an d oxygen. For service-aged tra nsform ers ra ted less tha n 69 kV, results of up to 35 ppm a re consider ed accepta ble. For 69 kV th rough 288 kV, the DGA test resu lt of 25 ppm is consider ed accepta ble. For great er th an 288 kV, moistur e should not exceed 20 ppm. However, the use of absolute values for wa ter does not a lways guara ntee sa fe conditions, an d the per cent by dry weight should be determ ined. See ta ble 12, “Doble Limits for I n-Service Oils,” in section 4.6.5. If values a re h igher, t he oil should be processed. If the tr an sform er is kept a s dry an d free of oxygen as possible, tra nsform er life will be exten ded. Reclama tion specifies that ma nufacturer s dry new tr an sform ers to no more th an 0.5% M/DW dur ing commissioning. In a t ra nsform er ha ving 10,000 poun ds of pap er insula tion, this m ean s th at 10,000 x 0.005 = 50 pounds of wat er (about 6 gallons) is in the pa per. This is not enough moisture to be detrimen ta l to electr ical integrity. When the tr an sform er is new, this water is distr ibut ed equa lly th rough the tra nsform er. It is extr emely importa nt to remove as m uch water as possible.
53
Table 10.—Typical Faults in Power Transformers [12] Fault Partial discharges
Examples Discharges in gas-filled cavities in insulation, resulting from incomplete impregnation, high moisture in paper, gas in oil supersaturation or cavitation, (gas bubbles in oil) leading to X wax formation on paper.
Discharges of low energy
Sparking or arcing between bad connections of different floating potential, from shielding rings, toroids, adjacent discs or conductors of different windings, broken brazing, closed loops in the core. Additional core grounds. Discharges between clamping parts, bushing and tank, high voltage and ground, within windings. Tracking in wood blocks, glue of insulating beam, winding spacers. Dielectric breakdown of oil, load tap changer breaking contact.
Discharges of high energy
Flashover, tracking or arcing of high local energy or with power follow-through. Short circuits between low voltage and ground, connectors, windings, bushings, and tank, windings and core, copper bus and tank, in oil duct. Closed loops between two adjacent conductors around the main magnetic flux, insulated bolts of core, metal rings holding core legs.
Overheating less than 300 °C
Overloading the transformer in emergency situations. Blocked or restricted oil flow in windings. Other cooling problem, pumps valves, etc. See the “Cooling” section in this document. Stray flux in damping beams of yoke.
Overheating 300 to 700 °C
Defective contacts at bolted connections (especially busbar), contacts within tap changer, connections between cable and draw-rod of bushings. Circulating currents between yoke clamps and bolts, clamps and laminations, in ground wiring, bad welds or clamps in magnetic shields. Abraded insulation between adjacent parallel conductors in windings.
Overheating over 700 °C
Large circulating currents in tank and core. Minor currents in tank walls created by high uncompensated magnetic field. Shorted core laminations.
Notes: 1. X wax formation comes from Paraffinic oils (paraffin based). These are not used in transformers at present in the United States but are predominate in Europe. 2. The last overheating problem in the table says �over 700 °C.” Recent laboratory discoveries have found that acetylene can be produced in trace amounts at 500 °C, which is not reflected in this table. We have several transformers that show trace amounts of acetylene that are probably not active arcing but are the result of hightemperature thermal faults as in the example. It may also be the result of one arc, due to a nearby lightning strike or voltage surge. 3. A bad connection at the bottom of a bushing can be confirmed by comparing infrared scans of the top of the bushing with a sister bushing. When loaded, heat from a poor connection at the bottom will migrate to the top of the bushing, which will display a markedly higher temperature. If the top connection is checked and found tight, the problem is probably a bad connection at the bottom of the bushing.
54
When th e tr an sform er is energized, water begins to migrate t o the coolest part of the tra nsform er a nd t he site of th e greatest electr ical str ess. This location is norma lly the insula tion in th e lower one-th ird of th e winding [5]. Pa per insu lat ion ha s a much great er affinity for wat er th an does the oil. The water will distr ibute itself un equa lly, with much more water being in the paper t ha n in the oil. The paper will partially dry th e oil by absorbing wat er out of th e oil. Temper at ur e is also a big factor in how the wat er distr ibutes itself between th e oil an d paper . See ta ble 11 below for compa rison. Table 11.—Comparison of Water Distribution in Oi l and Paper [5] Temperature (degrees C)
Water in Oil
Water in Paper
20°
1
3,000 times what is in the oil
40°
1
1,000 times what is in the oil
60°
1
300 times what is in the oil
The table above shows the tr emendous att ra ction th at pa per insulation has for wat er. The ppm of wat er in oil shown in t he DGA is only a sm all part of the wat er in t he tra nsform er. It is importa nt t ha t, when an oil sam ple is tak en, you record th e oil temper at ur e from the top oil temper at ure gage. Some labora tories give percent M/DW of th e insula tion in the DGA. Oth ers give percent oil sat ur at ion, an d some give only the ppm of wat er in th e oil. If you ha ve an accur at e temper at ur e of the oil and t he ppm of wat er, the N omogra ph (figur e 23, section 4.5.2) will give percent M/DW of the ins ulat ion a nd t he per cent oil sat ur at ion. Wh e r e d o e s t h e w a t e r c o m e f r o m ? Moisture can be in the insu lation when it is delivered from the factory. If th e tra nsform er is opened for inspection, th e insulat ion can a bsorb moistur e from the at mosphere. If ther e is a leak, moisture can enter in t he form of wat er or hum idity in a ir. Moistu re is also form ed by th e degrada tion of insulation a s the t ra nsform er a ges. Most wat er penetr at ion is flow of wet air or ra in water th rough poor gasket sea ls due t o pressure difference cau sed by tra nsform er cooling. Dur ing ra in or sn ow, if a t ra nsform er is rem oved from ser vice, some tr an sform er designs cool ra pidly an d the pressu re inside drops. The most comm on moistur e ingress points ar e gasket s between bush ing bottoms an d the t ra nsform er top an d th e pressu re r elief device gasket . Sm all oil leak s, especially in th e oil cooling piping, will also allow moistur e ingr ess. With ra pid cooling an d the resu ltan t pressu re drop, relatively lar ge amounts of water an d water vapor can be pum ped into the tra nsform er in a short time. It is importa nt t o repair sm all oil leaks; the sm all am ount of visible oil is not import an t in its elf, but it a lso indicat es a point wh ere moistur e will ent er [22].
It is critical for life extension to keep tr an sform ers a s dry a nd a s free of oxygen a s possible. Moisture a nd oxygen cause th e paper insu lation t o decay mu ch faster t ha n norma l and form acids, sludge, and more moistur e. Sludge settles on windings and
55
inside th e str uctu re, cau sing tr an sform er cooling to be less efficient, a nd s lowly over time temp era tu re rises. (This was discussed ear lier in “3. Tra nsform er Cooling Meth ods.”) Acids cau se an increase in th e ra te of decay, which form s more a cid, sludge, and m oistu re a t a faster ra te [20]. This is a vicious cycle of increasin g speed form ing more acid and causing more decay. The an swer is to keep the tr an sform er as dry a s possible and a s free of oxygen a s possible. In a ddition, oxygen inh ibitor sh ould be watched in t he DGA testing. The tr an sform er oil should be dried when m oisture rea ches the valu es accordin g to ta ble 12. Inh ibitor sh ould be added (0.3% by weight ASTM D-3787) when the oil is processed. Water can exist in a tr an sform er in five form s. 1. Fr ee water , at th e bott om of the ta nk. 2. Ice at t he t an k bottom (if the oil specific gra vity is grea ter th an 0.9, ice can float). 3. Water can be in the form of a wat er/oil emu lsion. 4. Water can be dissolved in t he oil an d is given in ppm in th e DGA. 5. Water can be in the form of hum idity if tra nsform ers ha ve an iner t gas blan ket. Fr ee water cau ses few problems with dielectr ic stren gth of oil; however, it should be drained as soon as possible. Ha ving a wat eroil int er face allows oil to dissolve water and t ran sport it to the insula tion. P roblems with moistur e in insulat ion were discussed above. If th e tr an sform er is out of service in wint er, wa ter can free ze. If oil specific gra vity is great er t ha n 0.9 (ice specific gra vity), ice will float . This can cau se tra nsform er failur e if the tra nsform er is ener gized with float ing ice ins ide. This is one reason th at DGA labora tories test specific gravity of transformer oil. The amount of moistur e tha t can be dissolved in oil increa ses with tem pera tu re. (See figur e 19.) This is why hot oil is used t o dryout a tr an sform er. A wat er/oil emu lsion can be form ed by pu rifying oil at too high temperat ure. When the oil
Figu re 19.—Maximum Amount of Wate r D i s s o l v e d i n M i n e r a l O i l Ve r s u s Temperature.
56
cools, dissolved moistu re form s an emu lsion [20]. A wat er/oil emulsion cau ses dr ast ic reduction in dielectric strength. How much moistur e in insulat ion is too mu ch? When th e insulation gets to 2.5% M/DW or 30% oil satu ra tion (given on some DGAs), the t ra nsform er sh ould have a d ry out with vacuum if the ta nk is ra ted for vacuum . If the tra nsform er is old, pulling a vacuum can do more ha rm th an good. In t his case, it is bett er to do roun d-the-clock re circulat ion with a Bowser dr ying the oil as m uch a s possible, which will pull water out of the paper. At 2.5% M/DW, th e paper insu lation is degrading much faster t ha n norm al [5]. As th e paper is degraded, more water is produced from th e decay products, and t he tr an sform er becomes even wett er an d decays even faster. When a tr an sform er gets a bove 4% M/DW, it is in dan ger of flash over if th e tem pera tu re r ises to 90 °C. 4.5.1 Diss olved Moisture in Transform er Oil. Moistu re is given in th e dissolved gas a na lysis in ppm, a nd some laborat ories also give percent sat ura tion. Percent sat ur ation means percent satu ra tion of water in th e oil. This is a percentage of how mu ch wa ter is in the oil compa red with t he m aximum am ount of wat er th e oil can hold. In figure 19, it can be seen th at the a mount of water th e oil can dissolve is grea tly dependen t on tem pera tu re. The curves (figure 20) below ar e percent sat ur at ion curves. On th e left line, find the ppm of wat er from your DGA. Fr om th is point, draw a horizont al with a stra ight edge. Fr om th e oil temper at ur e, dra w a vert ical line. At th e point where th e lines inter sect, read th e percent satu ra tion curve. If the point falls between two satu ra tion curves,
F i g u r e 2 0 .—T ra n s f o r m e r O i l P e r c e n t S a t u r a t i o n C u r v e s .
57
estimat e the percent sat ur at ion based on where the point is locat ed. For example, if the wat er is 30 ppm an d the t empera tu re is 40 °C, you can see on th e curves tha t th is point of int ersection falls about h alfway between t he 20% cur ve and t he 30% cur ve. This means t ha t th e oil is approximat ely 25% sat ura ted. Curves shown on figur e 20 ar e from I EE E 62-1995 [19]. C a u t i o n : B e l o w 3 0 °C , t h e c u r v e s a r e n o t v e r y a c c u r a t e . 4.5.2 Moisture in Transforme r Insulatio n. The illustra tion at right (figur e 21) shows how moistur e is distributed thr oughout t ra nsform er insu lation. Notice tha t t he m oisture is distributed a ccording to tempera tu re, with most moistur e at t he bott om a nd less as tempera tur e increases toward the top. In this example, there is almost twice th e moisture nea r bott om as th ere is at the top. Most service-aged t ra nsform ers fail in th e lower one-third of the windings, which is th e ar ea of most moisture. It is also the area of most electrical stress. Moistu re a nd oxygen ar e two of th e t ra nsform er’s worst enemies. It is very import an t to keep th e insulation a nd oil as dry a s possible and a s free of oxygen as possible. Failures due to moisture are the most common cause of t r a n s f o r m e r f a i l u r e s [5]. Without an accur at e oil tem pera tu re, it is impossible for laboratories to provide accurate inform at ion a bout th e M/DW or percent satur at ion. It will also be imp ossible for you to calculate this information accurately. Figu re 21.—Wate r Distribut ion in Experts disagree on how to tell T ra n s f o r m e r I n s u l a t i o n . how much moistur e is in t he insu lat ion based on how much moistur e is in th e oil (ppm). At best, met hods to determ ine moistu re in the insulation based solely on DGA are inaccur ate. The met hods discussed below to determine m oisture in the insu lation a re app roxima tions an d no decision sh ould be mad e based on one DGA. However,
58
keep in mind th at the life of the tra nsform er is the life of the insulation. The insu lat ion is quickly degra ded by excess moistur e an d th e presen ce of oxygen. Base a ny decisions on severa l DGAs over a period of time an d esta blish a t ren d of increasing m oisture. If the la b does not pr ovide th e percent M/DW, IEEE 62-1995 [19] gives a met hod. Fr om th e curve (figure 22), find tem perat ur e of the b o t t o m o i l s a m p l e a n d a d d 5 °C . Do not use th e top oil temper at ur e. This approxima tes tempera tur e of the bott om t hird (coolest pa rt ) of th e winding, where m ost of the wat er is located. Fr om t his tempera tur e, move up vertically to the cur ve. Fr om t his point on the curve, move horizont ally to the left an d find th e Myers Mult iplier nu mber . Tak e th is number a nd m ultiply the ppm of water shown on the DGA. The result is percent M/DW in t he upper par t of the insu lation. This meth od gives less amount of wat er t ha n t he Gener al Electric nomogra ph on th e following page.
Figu re 22.—Myers Multiplie r Versu s Tem perat ure .
59
This nomogra ph, published by Genera l Electric in 1974 (figure 23), gives the percent sat ur at ion of oil and percent M/DW of insulation. Use the nomograph to check your self after you h ave completed t he m ethod illustra ted in figure 22. The nomogra ph in figur e 23 will show more moistur e than the IEEE m ethod. The cur ves in figure 23 ar e useful to help understand relationships between temperatu re, percent saturation of the oil, an d percent M/DW of th e insulation. For exam ple, pick a point on th e ppm water line, say 10 ppm. Place a str aight edge on th at point an d pick a point on t he temperat ure line, say 45 °C. Read the percent satu rat ion a nd percent M/DW on t he cent er lines. In this example, percent sat ur at ion is about 6.5% an d th e % M/DW is a bout 1.5%. Now, hold th e 10 ppm point a nd move the sa mple t e m p er a t u r e u p wa r d (cooler), a nd notice h ow quickly the moistur e nu mbers increase. For exam ple, use 20 °C an d read t he % satu ra tion of oil at a bout 18.5% an d
Figu re 23.—Wate r Conten t of Pape r and Oil Nonogram.
60
th e % M/DW at a bout 3.75%. The cooler t he oil, th e higher t he m oistu re percentage for the sam e ppm of water in th e oil. Do not ma ke a decision on d ryout ba sed on only one DGA an d one calcula tion; it should be based on tren ds over a period of tim e. Tak e additional sam ples and send th em for a na lysis. Ta k e e x t r a c a r e t o m a k e s u r e t h e o i l t e m p e r a t u r e i s c o r r e c t . You can see by the nomogra ph t ha t m oisture cont ent var ies dram atically with temperatu re. Take e x t r a c a r e th at th e sample is not exposed to air. If after using the more conservat ive IEEE m ethod and again subsequent sam ples show M/DW is 2.5% or more an d th e oil is 30% sat ur at ed or m ore, the tr an sform er should be dried as soon a s possible. Check th e nomograph an d curves above to determine th e percent sat ur ation of th e oil. The insulation is degrading mu ch fast er tha n norma l due to the high moistur e conten t. Drying can be an expensive process; it is prud ent to consu lt with other s before ma king a final decision t o do dryout. However, it is mu ch less expensive to perform a dr yout th an t o allow a tra nsform er to degrade faster tha n norma l, substa ntially shortening t ra nsform er life.
4 .6 T ra n s f o r m e r O i l Te s t s T h a t S h o u l d B e D o n e A n n u a l l y Wi t h t h e D i s s o l v e d Gas Analys is. 4 .6 .1 D i e l e c t r i c S t r e n g t h . This test m easur es the voltage at wh ich t he oil electr ically breaks down. The test gives a good indicat ion of th e am ount of cont am ina nt s (wat er an d oxidation part icles) in the oil. DGA laborat ories typically use ASTM Test Method No. D-877 or D-1816. Th e a c c e p t a b le m i n i u m b r e a k d o w n v o l t a g e i s 3 0 k V fo r t r a n s f o r m e r s 2 8 7 .5 k V a n d a b o v e , a n d 2 5 k V fo r h i g h v o l t a g e t r a n s f o r m e r s r a t e d u n d e r 2 8 7 .5 k V . If the dielectric str ength t est falls below these nu mber s, the oil should be reclaimed. Do not base an y decision on one test resu lt, or on one type of test ; inst ead, look at a ll the information over several DGAs and establish trends before making any decision. The dielectric strength test is not ex tremely valuable; moisture in c o m b i n a t io n w i t h o x y g e n a n d h e a t w i l l d e s t r o y c e l lu l o s e i n s u l a t io n l o n g b e f o re t h e d i e l e c t ri c s t r e n g t h o f t h e o i l h a s g i v e n a c l u e t h a t a n y t h i n g i s g o i n g w r o n g [ 5 ]. The dielectric str ength test also reveals n othing a bout acids an d sludge. The tests explained below are mu ch more import an t. 4.6.2 Interfaci al Ten sion (IFT). This t est (ASTM D-791-91) [21], is used by DGA labora tories to determine t he int erfacial tension between the oil sample an d distilled wat er. The oil sam ple is put into a beaker of distilled wat er at a tem pera tu re of 25 °C. The oil should float because it s specific gra vity is less th an th at of wat er, which is one. There should be a distinct line between the t wo liquids. The IFT nu mber is th e amoun t of force (dynes) requ ired to pull a sma ll wire ring upward a distance of 1 centimeter thr ough th e water/oil interface. (A dyne is a ver y sma ll un it of force equa l to 0.000002247 poun d.) Good clea n oil will ma ke a very distinct line on top of the wat er a nd give an IFT n um ber of 40 to 50 dynes per centim eter of tr avel of th e wire ring.
61
As the oil ages, it is cont am inat ed by tiny pa rt icles (oxidation pr oducts) of th e oil an d paper insulation. These part icles extend a cross th e water /oil inter face line an d weaken th e tension between the t wo liquids. The more part icles, the weaker th e inter facial tension an d the lower the IFT num ber. The IFT and acid num bers together ar e an excellent indication of when th e oil needs to be reclaim ed. It is recomm ended th e oil be reclaimed when th e IFT n um ber falls to 25 dynes per cent imeter. At t his level, th e oil is very conta mina ted an d mu st be reclaimed to preven t sludging, which begins aroun d 22 dynes per cent imet er. See FIST 3-5 [20]. If oil is not r eclaim ed, sludge will sett le on wind ings, insula tion, etc., an d cause loadin g and cooling problems discussed in an ear lier section. This will grea tly shorten t ra nsform er life. There is a definite relationship between the acid num ber, the IFT, and the nu mber of year s in service. The a ccompan ying cur ve (figure 24) sh ows t he relationship and is found in many publications. (It was origina lly published in t he AIEE tr an sactions in 1955.) Notice th at the cur ve shows the normal service limits both for t he IFT a nd t he acid number. Figu re 24.—Interfacia l Tensio n, Acid 4.6.3 Acid Num ber. Acid N u m b e r , Ye a r s i n S e r v i c e . number (acidity) is the am ount of pota ssium h ydroxide (KOH) in milligra ms (mg) that it ta kes to neut ra lize the acid in 1 gram (gm) of tran sform er oil. The higher th e acid nu mber, th e more acid is in th e oil. New tr an sform er oils conta in pra ctically no acid. Oxidat ion of the insu lation a nd oils form s acids as th e tr an sform er a ges. The oxidat ion pr oducts form sludge and precipitate out inside the tr an sform er. The acids at ta ck met als inside the tan k an d form soaps (more sludge). Acid also at ta cks cellulose an d accelera tes insulat ion degra dation. Sludging ha s been found to begin wh en t he a cid nu mber rea ches 0.40; it is obvious t ha t t he oil should be reclaimed before it rea ches 0.40. I t is r e c o m m e n d e d t h a t t h e o i l be r e c l a i m e d w he n it rea che s 0.20 mg KOH/gm [20]. As with a ll oth ers, th is decision m ust not be based on one DGA test, but wa tch for r ising tr end in t he a cid num ber each year. Plan a head a nd begin budget planning before th e acid num ber reaches 0.20. 4 .6 .4 T e s t f o r Ox y g e n I n h i b i t o r Ev e r y 3 t o 5 Ye a r s w i t h t h e A n n u a l DGA Test . In pr evious sections, th e need to keep the t ra nsform er dr y and O 2 free was emp ha sized. Moistu re is destru ctive to cellulose and even more so in the
62
pres ence of oxygen. Some publicat ions st at e tha t each time you double th e moistur e (ppm), you halve the life of th e tra nsform er. As was discussed, acids ar e form ed tha t a tta ck th e insulation a nd m etals which form more acids, cau sing a viscous cycle. Oxygen inh ibitor is a key to exten ding the life of tr an sform ers. The inhibitor cur ren tly used is Diter tiar y But yl Pa ra cresol (DBPC). This work s sort of like a sacrificial anode in groun ding circuits. The oxygen att acks th e inhibitor instead of the cellulose insulation. As th is occurs an d th e tr an sform er a ges, th e inhibitor is used up an d needs to be repla ced. The ideal am ount of DBPC is 0.3% by tota l weight of th e oil (ASTM D-3487). Ha ve th e inh ibitor cont ent t ested with t he DGA every 3 to 5 years . If th e inh ibitor is 0.08% the transformer is considered uninhibited, and the oxygen freely attacks th e cellulose. If th e inhibitor falls to 0.1%, the tr an sform er should be re-inh ibited. For exa mple, if your tr an sform er t ested 0.1%, you n eed to go to 0.3% by adding 0.2% of th e total weight of th e tra nsform er oil. The na mepla te gives th e weight of oil—say 5,000 pounds—so 5,000 pounds X 0.002 = 10 pounds of DBPC needs to be add ed. It’s ok if you get a little t oo mu ch DBPC; this does not hur t t he oil. Dissolve 10 poun ds of DBPC in t ra nsform er oil th at you h ave heat ed to the sam e temper atu re as the oil inside the tr an sform er. It ma y ta ke some experimenta tion to get the r ight am oun t of oil to dissolve th e DBPC. Mix th e oil an d inhibitor in a clean cont ainer un til all th e DBPC is dissolved. Add th is mixtu re to the tr an sform er using the m ethod given in th e tra nsform er instr uction ma nua l for adding oil. C a u t i o n : Do not att empt th is unless you ha ve ha d experience. Conta ct an experienced cont ra ctor or experienced Reclam at ion people if you need help.
In either case, do not n eglect t his importan t m ainten an ce function; it is critical to tr an sform er insu lation t o have the proper am ount of oxygen inhibitor. 4 .6 .5 P o w e r F a c t o r . Power factor indicat es th e dielectr ic loss (leak age curr ent ) of th e oil. This test m ay be done by the DGA laborat ories. It m ay also be done by Doble test ing. A high power factor indicates deter iora tion and/or cont am ina tion by-products su ch as wat er, carbon, or other condu cting par ticles; met al soaps caused by acids (form ed as ment ioned a bove); att acking tra nsform er m etals; and products of oxidation. The DGA labs norm ally test th e power factor a t 25 °C an d 100 °C. Doble inform at ion [23] indicat es th e in-service limit for power fa ctor is less than 0.5% at 25 °C. If the power factor is greater tha n 0.5% an d less than 1.0%, further investigation is required; the oil may require replacement or fullers eart h filtering. If the pow er facto r is gre ate r than 1.0% at 25 °C, the oil may c a u s e f a i lu r e o f t h e t r a n s f o r m e r ; r e p l a c e m e n t o r r e c l a i m i n g i s r e q u i r e d. Above 2%, th e oil should be r emoved from s ervice an d r eclaimed or repla ced because equipment failure is a high probability. 4.6.6 Furans. Fu ra ns a re a fam ily of orga nic compoun ds which ar e formed by degra dat ion of paper insula tion (ASTM D-5837). Overhea ting, oxidation, and degradation by high moisture content contribute to the destruction of insulation
63
an d form furan ic compounds. Chan ges in furan s between DGA tests a re more importa nt t ha n individual num bers. The same is true for dissolved gases. Tran sform ers with greater th an 250 part s per billion (ppb) should be investigat ed becau se paper insu lat ion is being degra ded. Also look at t he IF T and acid number. Doble in-service limits a re r eproduced below to support th e above recomm ended guidelines. Table 12 below is excerpt ed from Doble En gineering Compa ny’s Reference Book on Insulating L iquids and Gases [23]. These Doble Oil Limit ta bles support inform at ion given in prior sections in th is FIST ma nua l and a re shown here a s summa ry tables.
Table 12.—Doble Limits for In-Service Oils Voltage Class 69 kV
>69
288 kV
>288 kV 1
Dielectric Breakdown Voltage, D 877, kV min
26
30
Dielectric Breakdown Voltage D 1816, .04-inch gap, kV, min.
20
20
25
Power Factor at 25 °C, D 924, max.
0.5
0.5
0.5
Water Content, D 1533, ppm, max.
2
2
2
Interfacial Tension, D 971, dynes/cm, min.
25
25
25
Neutralization Number, D 974, mg KOH/gm, max.
0.2
0.15
0.15
Visual Exam Soluble Sludge 1
35
25
clear and bright
clear and bright
3
3
ND
ND
20
clear 3
ND
D 877 test is not as sensitive to dissolved water as the D 1816 test and should not be used with oils for extra high voltage (EHV) equipment. Dielectric breakdown tests do not replace specific tests for water content. 2 The use of absolute values of water-in-oil (ppm) do not always guarantee safe conditions in electrical apparatus. The percent by dry weight should be determined from the curves provided. See the information in section. “4.5 Moisture Problems.” 3 ND = None detectable. These recommended limits for in-service oils are not intended to be used as absolute requirements for removing oil from service but to provide guidelines to aid in determining when remedial action is most beneficial. Remedial action will vary depending upon the test results. Reconditioning of oil, that is, particulate removal (filtration) and drying, may be required if the dielectric breakdown voltage or water content do not meet these limits. Reclamation (clay filtration) or replacement of the oil may be required if test values for power factor, interfacial tension, neutralization number, or soluble sludge do not meet recommended limits.
64
4.6.7 Taking Oil Sample s for DGA. Sam pling procedures a nd lab han dling ar e usua lly ar eas tha t cause the most problems in gett ing an a ccur at e DGA. There ar e times when atm ospheric gases, moistur e, or h ydrogen ta ke a sudden leap from one DGA to the next. As ha s been ment ioned, at t hese tim es, one should imm ediat ely ta ke an oth er sa mple to confirm DGA values. It is, of cour se, possible th at t he tr an sform er ha s developed an at mospheric leak, or th at a fau lt has sudd enly occur red inside. More often, the sam ple has not been tak en properly, or it ha s been cont am inated with a tm ospheric gases or mishan dled in other wa ys. The sa mple mu st be pr otected from a ll cont am ination, including at mospheric exposure.
Do not ta ke sam ples from th e sma ll sample port s locat ed on the side of the lar ge sam ple (dra in) valves. These port s ar e too sma ll to adequat ely flush th e large valve an d pipe nipple connected to the t an k; in a ddition, air can be drawn past the th reads an d conta mina te the sam ple. Fluid in th e valve an d pipe nipple remain dorm an t du ring opera tion a nd can be conta mina ted with moistur e, microscopic stem packing par ticles, and oth er pa rt icles. The volume of oil in t his location can also be cont am ina ted with gases, especially hydrogen. Hydr ogen is one of th e easiest gasses to form . With h ot sun on the side of th e tran sform er ta nk where th e sample valve is locat ed, high a mbient tem perat ur e, high oil tempera tur e, and captu red oil in th e sam ple valve an d extension, hydrogen form ed will stay in th is area until a sam ple is drawn. The lar ge sample (drain) valve can also be conta mina ted with hydrogen by galvan ic action of dissimilar m etals. Sam ple valves are usu ally brass, an d a br ass pipe plug should be inst alled when t he valve is not being used. If a galvanized or black iron pipe is installed in a br ass valve, the dissimilar meta ls produce a th erm ocouple effect, and circulat ing curr ent s are produced. As a result , hydr ogen is gener at ed in the void between t he plug and valve gat e. If th e valve is not flushed v e r y th oroughly th e DGA will show high h ydrogen. Oil should not be sam pled for DGA pur poses when th e tr an sform er is at or below freezing temperat ur e. Test values which a re a ffected by water (such as dielectr ic str ength , power factor, an d dissolved moistur e content ) will be inaccura te. C a u t i o n : T ra n s f o r m e r s m u s t n o t b e s a m p l e d i f t h e r e i s a n e g a t i v e pressure (vacuum) at the sample valve.
This is typically not a pr oblem with conservat or t ra nsform ers. If the tra nsform er is nitrogen blanketed, look at the pr essure/vacuum gage. If the pressur e is positive, go ah ead an d tak e the sam ple. If th e pressure is negative, a vacuum exists at th e top of th e tra nsform er. If ther e is a vacuum at t he bottom, air will be pulled in when th e sample valve is opened. Wait unt il the pr essure gage reads positive before sam pling. P u l l i n g i n a v o l u m e o f a i r c o u l d b e d i s a s t r o u s i f the transformer is ene rgized. If negat ive pressur e (vacuu m) is not t oo high, the weight of oil (hea d) will mak e positive pressur e at the sa mple valve, and it will be safe to take a sam ple. Oil
66
hea d is about 2.9 feet (2 feet 10.8 inches) of oil per pounds p er squ ar e inch (psi). If it is importa nt to take t he sam ple even with a vacuum showing at t he top, proceed as described below. Use th e sam ple tubing and ada ptors described below to adapt t he large sam ple valve to -inch t ygon t ubing. Fill a length (2 to 3 feet) of tygon tubin g with new tr an sform er oil (no air bubbles) an d at ta ch one end t o the pipe plug an d th e oth er end to the sma ll valve. Open the large sam ple (dra in) valve a sma ll am ount an d very slowly cra ck open t he s ma ll valve. I f o i l i n t h e t y g o n t u b i n g m o v e s t o w a r d t h e t r a n s f o r m e r , s h u t o f f t h e v a l v e s i m m e d i a t e l y . D o n o t a l lo w a i r t o b e p u l l e d i n t o t h e t r a n s f o r m e r . If oil moves toward the tr an sform er, th ere is a vacuu m a t th e sample valve. Wait unt il th e pressure is positive before ta king th e DGA sam ple. If oil is push ed out of th e tygon tu bing int o th e waste cont ain er, th ere is a positive pressu re an d it is safe to proceed with DGA sam pling. Shu t off th e valves and configure th e tubing and valves to take th e sample per t he inst ru ctions below. D G A O i l S a m p l e C o n t a i n e r . Glass sample syringes are recomm ended. There ar e different conta iners such as sta inless steel vacuum bottles and other s. It is recomm ended tha t only glass syringes be used. If ther e is a sm all leak in th e sampling tubing or connections, vacuum bottles will draw air into the sample, which can not be seen inside the bott le. The sam ple will show high at mospher ic gases and high moisture if the air is humid. Other conta mina tes such as suspended solids or free water can not be seen inside the vacuum bott le. Glass syringes are t he simplest t o use becau se air bubbles ar e easily seen a nd expelled. Other cont am inates a re easily seen, and an other sam ple can be immediat ely ta ken if the sa mple is cont am inated. The downside is tha t glass syringes must be ha ndled car efully an d mu st be pr otected from direct su nlight. They should be retu rned to their shipping conta iner immediately after tak ing a sam ple. If they ar e exposed to sunlight for a ny tim e, hydrogen will be genera ted a nd t he DGA will show false hydr ogen rea dings.
For t hese rea sons, glass syringes ar e recomm ended, and the instru ctions below include only this sampling method. Obtain a brass pipe plug (norma lly 2 inches) that will threa d into the sa mple valve at the bottom of th e tra nsform er. Drill an d ta p th e pipe plug for -inch NPT and insert a -inch pipe nipple (brass if possible) an d at ta ch a sm all -inch valve for cont rolling th e sam ple flow. Atta ch a -inch t ygon t ubing ada ptor to th e sma ll valve outlet. Sizes of th e piping and thr eads above do not ma tter ; an y arr angement with a small sample valve and adaptor to -inch t ygon tu bing will su ffice. See figur e 25.
67
Taking the Samp le. •
Remove the existing pipe plug and inspect t he valve opening for r ust an d debris.
•
Crack open th e valve an d allow just enough oil to flow into th e wast e container to flush the valve an d th reads. Close the valve and wipe the th reads an d outlet with a clean dry cloth.
Figu re 25.—Oil Sam pling P iping .
•
Re-open t he valve slightly an d flush approxima tely 1 quart into the wa ste container.
•
Inst all th e bras s pipe plug (described above) an d associat ed -inch pipe an d sma ll valve, and a short piece of new -inch t ygon t ubing to th e out let of th e -inch valve.
•
Never use the sam e sample tubing on different tr an sform ers. This is one way a sa mple can be conta mina ted a nd give false readings.
•
Open both th e large valve an d sma ll sam ple valve and allow another quar t t o flush th rough the sam pling appar at us. Close both valves. Do this before at ta ching the glass sam ple syringe. Make sur e the sh ort piece of tygon t ubing tha t will att ach to the sam ple syringe is insta lled on t he -inch valve before you do this.
•
Insta ll the glass sam ple syringe on t he sh ort piece of -inch tu bing. Turn the stopcock han dle on t he syringe so th at the ha ndle points t oward the syringe. N o t e : T h e h a n d l e a l w a y s p o i n t s t o w a r d t h e c l o s e d p o r t . The oth er two ports ar e open to each oth er. See figure 26.
F i g u r e 2 6 .—S a m p l e S y r i n g e ( F l u s h i n g ) .
68
•
O pe n t h e la r g e sa m p l e v a lv e a s m a l l a m ou n t a n d a d j u st t h e -inch valve so th at a gent le flow goes th rough t he flush ing port of th e glass syr inge into the waste bucket.
•
Slowly tur n t he syringe stopcock ha ndle so that the h an dle points t o the flush ing port (figur e 27). This closes t he flush ing an d a llows oil to flow int o the sa mple syringe. Do not pull the syringe handle; th is will create a vacuum an d allow bubbles to form. The syringe ha ndle (piston) should back out ver y slowly. If it moves too fast , adjust t he small -inch valve unt il the syringe slows, an d h old your han d on t he back of th e piston so you can cont rol the Figu re 27.—Sam ple Syrin ge (Filling). travel.
•
Allow a sma ll amoun t, a bout 10 cubic centimet ers (cc), to flow int o the s yringe an d tur n th e stopcock ha ndle again so tha t it point s to the syringe. This will aga in a llow oil to come out of the flushin g port into th e wast e bucket.
•
Pu ll th e syringe off th e tu bing, but d o not sh u t off th e oil flow. Allow th e oil flow to continu e into the waste bucket.
•
Hold the syringe vertical an d tur n th e stopcock up so th at th e ha ndle points a way from the syringe. Pr ess the syringe piston to eject an y air bu bbles, but leave 1 or 2 cc oil in the syringe. See the accompa nying figure 28. C a u t i o n : Do not eject a ll the oil, or air will reenter. Figure 28.—Sample Syringe Bubble Removal.
•
Turn th e stopcock ha ndle toward t he syringe. The sma ll am ount of oil in the syringe should be free of bubbles and r eady to receive th e sam ple. If th ere ar e still bubbles at t he t op, repeat the process un til you h ave a sma ll amount of oil in th e syringe with no bubbles.
•
Reatt ach the tygon tu bing. This will again allow oil to flow out of the flushing port. Slowly tu rn t he stopcock ha ndle toward th e flush ing port which aga in will allow oil to fill th e syrin ge. The syr inge pist on will aga in ba ck slowly out of th e syringe. Allow th e syringe to fill abou t 80% full. Hold the pist on so you can st op its m ovemen t a t a bout 80% filled. C a u t i o n : Do not pull the piston. This will cau se bubbles to form .
69
•
Close the stopcock by tu rn ing the ha ndle toward th e syringe. Oil aga in will flow int o th e waste cont ainer . Shu t off both va lves, rem ove the sam pling appa ra tu s, and reinsta ll th e origina l pipe plug. C a u t i o n : Do not eject a ny bubbles th at form after th e sam ple is collected; these a re gases th at should be included in t he lab sam ple.
•
Retur n t he syringe to its original conta iner imm ediately. D o n o t a l l o w s u n l i g h t t o i m p a c t t h e c o n t a i n e r f o r a n y l e n g t h o f t i m e . Hydr ogen will form an d give false r eadin gs in th e DGA.
•
Carefully package the syringe in the sam e man ner th at it was shipped to the facility a nd sen d it t o the lab for processing.
•
Dispose of waste oil in the plant waste oil container.
4.6.8 Silicon e Oil-Filled Transform ers. Silicone oils becam e m ore comm on when PCBs were discontinu ed. They are mainly used in tr an sform ers inside buildings and tha t ar e smaller tha n generat or step-up tra nsform ers. Silicone oils ha ve a higher fire point t ha n m ineral oils and, th erefore, ar e used wher e fire concern s ar e more critical. As of th is writing, ther e are n o definitive published sta nda rds. IEE E ha s a guide and Doble ha s some service limits, but th ere are no sta nda rds. Inform at ion below is taken from th e IEE E pu blicat ion, from Doble, from ar ticles, from I EC 60599 concepts, a nd from Delta X Resear ch’s/Tran sform er Oil Ana lyst ru les. Silicone oil dissolved gas a na lysis is in th e beginn ing sta ges, an d th e suggested m ethods and limits below ar e subject t o cha nge as we gain more experience. However, in th e absence of an y oth er met hods and limits, use th e ones below as a beginn ing.
Silicone oils used in tr an sform ers a re polydimet hylsiloxane fluids, which a re different t ha n m ineral oils. Many of the gases generat ed by ther ma l and electr ical fau lts are th e same. The gases are genera ted in different pr oport ions th an with tr an sform er m inera l oils. Also, some fau lt gases ha ve differen t solubilities in silicone oils th an in min era l oils. Ther efore, th e sam e fau lts would produce different concentrations and different generation rates in silicone oils than mineral oils. As with m inera l oil-filled tr an sform ers, th ree pr incipal causes of gas gener at ion are aging, thermal faults, and/or electrical faults resulting in deterioration of solid insu lat ion an d deter iora tion of silicone fluid. These faults h ave been discussed a t length in p rior sections an d will not be discussed in great deta il her e. Overhea ting of silicone oils causes degra dat ion of fluid an d gener at ion of gases. Gases gener at ed depend on th e am ount of dissolved oxygen in t he fluid, temper atu re, and h ow close bare copper condu ctors ar e to the heat ing. When a tr an sform er is n ew, silicone oil will typically cont ain a lot of oxygen. Silicone tr an sform ers ar e typically sealed an d pressu rized with n itrogen. New silicone oil is not degass ed; and, a s a ru le, oxygen concent ra tion will be equivalent to oxygen solubility (ma ximum ) in silicone. The silicone h as been exposed to atm osphere for 70
some time dur ing manufacture of the tra nsformer and man ufactur er an d storage of silicone oil itself. The refore, car bon monoxide and car bon dioxide ar e easily form ed an d dissolved in th e silicone du e to th e abu nda nce of oxygen in th e oil resulting from t his atm ospheric exposur e. In n orm al new silicone tr an sform ers (no faults), both carbon m onoxide an d carbon dioxide will be genera ted in th e initial years of operat ion. As th e tra nsform er ages an d oxygen is depleted, genera tion of th ese gases slows and concent ra tions level off [25]. See figur e 29 below for th e r elationship of decrea sing oxygen an d increa sing carbon m onoxide an d car bon dioxide as a t ra nsform er ages. This cur ve is for gener al inform at ion only an d should not be taken to represent any par ticular t ra nsform er. A real tr an sform er with cha nges in loading, am bient t empera tur es, an d various dut y cycles would ma ke t hese curves look t ota lly differen t. After th e transformer is older (assuming n o fau lts have occurred), oxygen concent ra tion will reach equilibrium (figur e 29). Reaching equilibrium ma y ta ke a few years depending on t he size of th e transformer, loading, ambient temperatu res, etc. After th is time, oxygen, ca rbon monoxide, an d Figu re 29.—Relatio nsh ip of Oxyge n to Carbon carbon dioxide D i o x i d e a n d C a r b o n Mo n o x i d e a s T r a n s f o r m e r Ag e s . level off and the ra te of production of th ese gases from n orma l aging should be rela tively const an t. If generation rates of these gases change greatly (seen from the DGA), a fault has occurr ed, either t her ma l or electr ical. Rat e of genera tion of th ese gases a nd am ount s can be used to roughly determ ine what th e fau lt is. Once you notice an significan t increa se in ra te of genera tion of any gas, it is a good idea t o subtr act th e amount of gas tha t was alrea dy in th e tra nsform er before th is increase. This ensur es tha t gases used in the diagnosis are only gases tha t were generat ed after th e fau lt began. Carbon m onoxide will be a lot h igher in a silicone t ra nsform er t ha n a minera l oil-filled one. Th e difficulty is in t rying t o determ ine wha t is producing th e CO; is it comin g from n orm al a ging of oil or from deter iora tion of pap er from a fault condit ion. The only solut ion is a fura n an alysis. If th e CO cont ent is grea ter t ha n th e IEE E limit of 3,000 ppm [26], and t he gener at ion r at e G1 is met or exceeded, a fur an a na lysis is recomm ended with t he an nua l DGA. If a th erma l fau lt is
71
occur ring an d is producing CO and sm all amount s of meth an e an d hydrogen, the fault m ay be ma sked by th e norm al pr oduction of CO from t he silicone oil itself. If th e CO generat ion r at e ha s great ly increased, along with other gases, it becomes obvious th at a fau lt ha s occur red. The fura n an alysis can only tell you if th e paper is involved (being hea ted) in th e fault. Some gener al conclusions can be dra wn by compa ring silicone oil and miner al oil transformers. 1. All silicone oil filled tr an sform ers will ha ve a great dea l more CO tha n norm al min era l oil filled tr an sform ers. CO can come from two sour ces, th e oil itself and from degra dat ion of pap er insu lation. If th e DGA shows little oth er fault in gas genera tion besides CO, the only way to tell for cert ain if CO is coming from pa per degradat ion (a fau lt) is to ru n a fura n a nalysis with th e DGA. If other fau lt gases ar e also being gener at ed in significan t a moun ts, in ad dition t o CO, obviously th ere is a fault, a nd CO is coming from paper degradat ion. 2. Ther e will genera lly be more hydrogen presen t th an in a m iner al oil-filled transformer. 3. Due t o “fault m ask ing,” men tioned above, it is almost imp ossible to diagnose wha t is going on ins ide a silicone filled tr an sform er ba sed solely on DGA. One exception is if acetylene is being gener at ed, ther e is an a ctive ar c. You must also look a t gas generat ion r at es an d opera ting history. Look a t loading history, th rough fau lts, and oth er incidents. It is imperat ive tha t deta iled records of silicone oil filled tr an sform ers be carefully kept u p-to-date. These a re invalu able when a problem is encoun tered. 4. If acetylene is being generat ed in any a mount, th ere is a definitely an active electr ical ar c. The tr an sform er should be rem oved from service. 5. In gener al, oxygen in a silicone-filled tr an sform er comes from at mospher ic leaks or was present in t he tr an sform er oil when it was new. This oxygen is consum ed as CO an d CO 2 ar e form ed from the norma l heat ing from operation of the t ran sformer. 6. Once the t ra nsform er ha s ma tu red an d the oxygen ha s leveled off an d rem ained r elatively consta nt for t wo or more DGA sam ples, if you see a sudden increase in oxygen, and perh aps carbon dioxide an d nitr ogen, the tr an sform er h as developed a leak. In ta ble 14 below are IE EE limits [26], compar ed with Doble [25] in a stu dy of 299 opera ting tr an sform ers. The ta ble of gases from th e Doble stu dy seems more rea listic. They show gas level avera ge of 95% of tr an sform ers in th e study. Note, with t he last four gases, limits given by th e IEE E (trial u se guide) ru n over 70% higher th an t he Doble 95% norms. But with t he first t hr ee gases, hydrogen, meth an e, and eth an e, the IEE E limits ar e well below the a mount of gas found in 95% norm s in th e Doble study. We o b v i o u s l y c a n n o t h a v e l i m i t s t h a t a r e below the amoun t of gas found in normal operating transformers. 72
Ther efore, it is suggested t ha t we use th e Doble (95% norm ) limits. The 95% norm limit m eans th at 95% of the silicone oil tr an sform ers st udied ha d gas levels below th ese limit s. Obviously, 5% ha d gases higher t ha n th ese limits . These are problem tr an sform ers th at we should pay more at tent ion to.
Table 14.—Comparison of Gas Limits Gas
Doble 95% Norm
IEEE Limits
Hydrogen
511
200
Methane
134
100
Ethane
26
30
Ethylene
17
30
Acetylene
0.6
1
CO
1,749
3,000
CO2
15,485
30,000
Total Combustibles
2,024
3,360
In t able 15, the IEE E limits for L1 were chosen. For L2 limits, a sta tistical an alysis was applied, an d two stan dar d deviations were added to L1 to obtain L2. For L3 limits, th e L1 limit s were doubled. Table 15.—Suggested Levels of Concern (Limits) L1 (ppm)
L2 (ppm)
L3 (ppm)
G1 (ppm per month)
G2 (ppm per month)
Hydrogen
200
240
400
20
100
Methane
100
125
200
10
50
Ethane
30
40
60
3
15
Ethylene
30
25
60
3
15
Acetylene
1
2
3
1
1
CO
3,000
3,450
6,000
300
1,500
CO2
30,000
34,200
60,000
1,500
15,000
TDCG
3,360
3,882
6,723
na
na
Gas
Gas generat ion r at e limits G1 are 10% of L1 limits per month. G2 genera tion ra te limits ar e 50% of L1 limits per m onth . These basic concepts wer e tak en from IEC 60599 [12], for miner al oil tr an sform ers a nd a pplied to silicone oil tra nsform ers
73
due t o absen ce of an y oth er criter ia. As our experience grows in silicone DGA, th ese may h ave to be cha nged, but they will be used in the beginning. Limits L1, L2, an d L3 represent t he concentra tion in individual gases in ppm. G1 an d G2 represent s genera tion ra tes of individual gases in ppm per month. To obtain G1 an d G2 in ppm per day divide the per month n um bers by 30. Except for a cetylene, G1 is 10% of L1 an d G2 is 50% of L1. The gen er at ion r at es (G1, G2), ar e point s wher e our level of concern should increa se, especially when considered with th e L1, L2, an d L3 limits. At G2 generat ion r at e, we should be extremely concerned a nd reduce th e DGA sampling interval accordingly, an d perh aps plan an outa ge, etc. Except for a cetylene, genera tion ra te levels G1 and G2 were t ak en from IEC 60599 referen ce [12] which is used with m inera l oil tr an sform ers. An y a m o u n t o f o n g o i n g a c e t y l e n e g e n e r a t io n m e a n s a c t i v e a r c i n g i n s i d e t h e t r a n s f o r m e r . I n th i s c a s e , t h e t r a n s f o rm e r s h o u l d b e r e m o v e d f r o m s e r v i c e . These criter ia were chosen becau se of an absen ce of an y oth er criter ia. As dissolved gas a na lysis criter ia for silicone oils becomes bett er kn own a nd qua nt ified table 15 will cha nge to reflect n ew inform at ion. A s w i t h m i n e r a l o i l -f i ll e d t r a n s f o r m e r s , g a s g e n e r a t i o n r a t e s a r e m u c h m o r e i m p o r t a n t t h a t t h e a m o u n t o f g a s p r e s e n t . Tota l accum ulated gas depends a lot on age (an older tra nsform er ha s more gas). If the ra te of generat ion of an y combust ible gas sh ows a su dden increa se in th e DGA, tak e an oth er oil sam ple imm ediately to confirm th e gas genera tion r at e increase. If the second DGA confirms a gen era tion ra te increas e, get some out side advice. Be car eful; gas generat ion ra tes increase somewhat with tem perat ur e var iations caused by increased loading and summ er ambient temper at ur es. However, higher operat ing temper atu res a re a lso the m ost likely conditions for a fau lt to occur . The rea l question is ha s th e increased gas generat ion r at e been cau sed by a fault or increased temper at ure from greater loading or higher ambient tem perat ur e?
If gas genera tion r at es ar e fairly consta nt (no big increases a nd less th an G1 limits above), what do we do if a t ra nsform er exceeds th e L1 limits? We begin to pay more att ention to tha t tr an sform er, just a s we do with a minera l oil tr an sform er. We may shorten th e DGA sam pling interval, reduce loading, check tr an sform er cooling, get some outside a dvice, etc. As with min era l oil tr an sform ers, age exert s a big influen ce in accum ula ted gas. We should be mu ch more concern ed if a 3-year old tr an sform er which ha s exceeded th e L1 limits t ha n if a 30-year old tra nsform er exceeds the limits. However, if G1 gener at ion ra tes ar e exceeded in either an old or n ew tra nsform er, we should step up our level of concern. If accum ulated gas exceeds the L2 limit, we ma y plan to ha ve the tr an sform er degassed. Exam ine the physical tests in the DGAs and compa re th em to the Doble/IEE E t able (table 16) ( Reference Book on Insu lating L iquids and Gasses ) [23]. The oil should be treat ed in whatever ma nner is appropriat e if th ese limits ar e exceeded.
74
If both L1 limit s a nd G1 limits ar e exceeded, we should become m o r e c o n c e r n e d . Reduce sam pling int erva ls, get out side advice, reduce loadin g, check tr an sform er cooling a nd oil levels, et c. I f G 2 g e n e r a t i o n l i m i t s a r e e x c e e d e d , w e s h o u l d b e e x t r e m e l y c o n c e r n e d . It will not be long before L3 limits a re exceeded, an d consider at ion m ust be given to removing the tr an sform er from service, for testin g, repair, or replacement . If acetylen e is being gene rated, the transformer should be taken ou t of s e r v i c e . However as with m ineral oil tra nsform ers, a one-time near by light ning str ike or t hr ough fau lt can caus e a “one-time” genera tion of acetylene. If you notice acetylene in t he DGA, immediately take an oth er sa mple. I f t h e a m o u n t o f a c e t y l e n e i s i n c r e a s i n g , a n a c t i v e e l e c t r i c a l a rc i s p re s e n t w i t h i n t h e transformer. It should be taken out of service .
If you h ave a critical silicone (or m inera l oil-filled tra nsform er), such a s a single sta tion service tr an sform er, or excita tion tr an sform er, you sh ould find out if a spar e is available at a nother facility or from Western Area Power Administra tion or Bonn eville Power. If th ere a re no oth er possible spar es consider beginn ing the budget process for getting a spare tra nsform er. Table 16 lists test limits for s ervice-aged silicone filled tr an sform er oil. If any of th ese limits a re exceeded, it is suggested th at the oil be trea ted in wha tever ma nner is appropriate t o retur n t he oil to serviceable condition.
Table 16.—Doble and IEEE Physical Test Limits for Service-Aged Silicone Fluid Unacceptable Values Indicate
ASTM Test Method
Test
Acceptable Limits
Visual
Clear free of particles
Particulates, free water
D 1524 D 2129
Dielectric breakdown voltage
30 kV
Particulates, dissolved water
D 877
Water content maximum
70 ppm (Doble) 100 ppm (IEEE)
Dissolved water contamination
D 1533
Power factor max. at 25 °C
0.2
Polar/ionic contamination
D 924
Viscosity at 25 °C, cSt
47.5–52.5
Fluid degradation contamination
D 44
Acid neutralization number max, mg KOH/gm
0.1 (Doble) 0.2 (IEEE)
Degradation of cellulose or contamination
D 974
Note: If only one number appears, both Doble and IEEE have the same limit.
75
If the a bove bove limits a re exceeded in t he DGA, th e silicone silicone oil oil should be filtered, dried or treated to correct the specific problem.
4.7 4.7 Transforme r Testing Wh e n t h e t r a n s f o r m e r i s n e w b e f o r e e n e r g i z i n g a n d e v e r y 3 t o 5 y e a r s , t h e t r a n s f o r m e r a n d b u s h i n g s s h o u l d b e D o b l e t e s t e d . Tran sfo sform er t esting falls falls into thr ee broad broad categories: categories: Fa ctory testing when when t he tr an sfo sform er is new or or ha s been been refurbished, acceptance testing upon delivery, and field testing for maintenance and diagnostic pur poses. Some tests at th e factory factory are comm comm on to most power power tr an sform sform ers, but ma ny of of th e factory factory tests ar e tra nsform nsform er- specif specific ic.. Table 17 lists severa severa l tests. This test cha cha rt ha s been adapted from from IE EE 62-1995 62-1995 reference reference [19] [19].. Not all of of the listed tests ar e done done a t t he factory, factory, and not all of of them ar e done in th e field. field. Ea ch t ra nsform nsform er a nd ea ch situ at ion ion is diff different , requiring its own own un ique approach approach and tests.
Deta ils of how to ru n specific specific test s will will not not be add ressed in t his FI ST. It would be impr actical to repea t h ow to do Doble Doble testing of a t ra nsform nsform er when th e informa informa tion is rea dily ava ilable in in Doble publicat publicat ions. ions. With some exceptions, exceptions, this is tr ue for for m ost of of the t ests. Specif Specific ic info inform at ion ion is readily available available with with in the t est instr um ent ma nu factur ers literat literat ur e. Another example is the tr an sfo sform er tur ns rat io test (TTR) (TTR);; specifi specificc test inform inform at ion ion is ava ilable with th e inst inst ru men t. However, inform inform at ion ion on on some test s m ay n ot be available a nd will be cov covered ered br iefly. iefly. field to check check 4.7. 4.7.1 1 Windin g Res istan ce s. Winding resista nces ar e test ed in th e field for loose connections, broken strands, and high contact resistance in tap changers. Key gases gases in creasing in th e DGA will will be eth an e an d/or d/or et hylene an d possibly possibly meth an e. Results are compa compa red to other other phases in wye co conn ected ected tr an sfo sform ers or or between pa irs of ter mina ls on on a d elta-conn elta-conn ected ected windin g to deter mine if a resist an ce is too too high. Resista nces can can also be be compar compar ed to th e original original factory factory mea sur emen ts. Agreem ent with in 5% for for an y of th e above above compa compa risons is considered satisfactory. satisfactory. You m ay ha ve to convert r esistance measur ement s to the reference temperature used at the factory (usually 75 °C) to compare your resistan ce measu remen ts t o the factory factory results. To do this use th e fo follo llowing formula:
Rs
= Rm
Ts + Tk Tm + Tk
Rs = Resista Resista nce at t he factory factory reference reference tempera tur e (fo (found in th e tr an sfo sform er manual) Rm = Resistance you you a ctu ally measured Ts = Fa ctory reference reference temper at ur e (usually 75 75 °C) °C) Tm = Tempera Tempera tur e at which which you you took took the mea sur ement s Tk = a consta consta nt for t he par ticular ticular meta l the winding is is ma de from: from: 234.5 °C for copper 225 °C for aluminum 76
It is very difficult to determine actual winding temperature in the field, and, normally, this is not needed. You only need to do the above temperature corrections if you are going to compare resistances to factory values. Normally, only the phase resistances are compared to each other, and you do not need the winding temperature to compare individual windings.
You can compare winding resistances to factory values; change in these values can reveal serious problems. A suggested method to obtain an accurate temperature is outlined below. If a transformer has just been de-energized for testing, the winding will be cooler on the bottom than the top, and the winding hot spot will be hotter than the top oil temperature. What is needed is the average winding temperature, and it is important to get the temperature as accurate as possible for comparisons. The most accurate method is to allow the transformer sit de-energized until temperatures are equalized. This test can reve al serious problems, so it’s worth the effort. Winding resistances are measured using a Wheatstone Bridge for values 1 ohm or above and using a micro-ohmmeter or Kelvin Bridge for values under 1 ohm. Multi-Amp (now AVO) makes a good instrument for these measurements which is quick and easy to use. Take readings from the top of each bushing to neutral for wye connected windings and across each pair of bushings for delta connected windings. If the neutral bushing is not available av ailable on wye connected windings, you can take each one to ground (if the neutral is grounded), or take readings between pairs of bushings as if it were a delta winding. Be consistent each time so that a proper comparison can be made. The tap changer can also be changed from contact to contact, and the contact resistance can be checked. Keep accurate records and connection diagrams so that later measurements can be compared. 4.7.2 Core Insulation Resistance and Inadvertent Core Ground Test. Core insulation resistance and inadvertent core ground test is used if an additional core ground is suspected; this may be indicated by the DGA. Key gases to look for are ethane and/or ethylene and possibly methane. These gases may also be present if there is a poor connection at the bottom of a bushing or a bad tap changer contact. Therefore, this test is only necessary if the winding resistance test above shows all the connections and if tap changer contacts are in good condition.
The intentional core ground must be disconnected. This may be difficult, and some oil may have to be drained to accomplish this. On some transformers, core grounds are brought outside through insulated bushings and are easily accessed. A standard dc megohmmeter is then attached between the core ground lead (or the top of the core itself ) and the tank (ground). The megohmmeter is used to place a dc voltage between these points, and the resistance measured. A new transformer should read greater than 1,000 megohms. A service-aged transformer should read greater than 100 megohms. Ten to one-hundred megohms is indicative of deteriorating insulation between the core and ground. Less than 10 megohms is 77
suffic sufficient ient t o cau cau se destru ctive ctive circ circulating ulating curr ents an d m ust be fur fur ther investigat ed [19]. [19]. A solid solid core core ground m ay r ead zer o ohm ohm s; this, of of course, course, causes destru ctive circ circulat ulat ing curren ts also. also. Some limited su ccess ccess h as been obta obta ined in “bur “bur ning off” off” un inten tional core core grounds grounds u sing a dc or a c cur rent sour sour ce. This is is a risky operat operat ion, ion, and t he curren t ma y cau cau se additiona additiona l dama ge. The current sour sour ce is norm norm ally limited limited to 40 to 50 amps ma ximu ximu m a nd sh ould be increased increased slowly slowly so so as to use as little current as possib possible le to acco accomplish the t ask. This should should only only be used as a last r esort esort an d th en only with consultation from the manufacturer, if possible, and with others experience experienced d in th is task.
78
Table 17.—Transformer Test Summary Chart Part to be Tested
Test to be Performed
W indings
Resistance Across W indings Turns Ratio/Polarity/Phase Excitation Current at All Tap Positions Short Circuit Impedance Insulation Resistance to Ground (megohmmeter) Capacitance (Doble) Power Factor/Dissipation Factor (Doble) Induced Voltage/Partial Discharge/Riv
Bushings
Capacitance (Doble) Dielectric Loss (Doble) Power Factor/Dissipation Factor (Doble) Partial Discharge (Doble) Temperature (Infrared) Oil Level (Sight Glass) Visual Inspection (Cracks and Cleanliness)
DGA Insulating Oil
Dissolved Gas Analysis Dielectric Strength Interfacial Tension Acid Number Visual Inspection Color Water Content Oxygen Inhibitor Power Factor/Dissipation Factor
Tap Changers - Load
Contact Pressure and Continuity Temperature (Infrared) Turns Ratio at All Positions Timing Motor Load Current Limit Switch Operation and Continuity
Tap Changers - No Load
Contact Pressure and Continuity Centering Turns Ratio at All Positions Visual Inspection
Core
Core Insulation Resistance to Tank Ground Test (megohmmeter)
Tanks an and As Associated De Devices
Pressure/Vacuum/Temperature Ga Gages - Calibration Temperature (Infrared) Visual Inspection (Leaks and Corrosion)
Conservator
Visual Inspection (Leaks and Corrosion)
Air Drier Desiccant
Proper Color Valves in Proper Position
Sudden Pressure Relay
Calibration and Continuity
Buchholz Relay
Proper Operation and Continuity
Cooling System
Temperature (Infrared)
Heat Exchanger Radiators
Clear Air Flow Visual (Leaks, Cleaning, and Corrosion)
Fans
Controls Visual Inspection and Unusual Noise
Pumps
Rotation and Flow Indicator Motor Load Current
79
REFERENCES
1.
IEEE Stan dard C57.12.01-1989 Stan dard General Requirements for Dry-Type Distribution, Power, and Regulating Transformers (ANSI).
2.
IEEE Stan dard C57.12.00-1993 Stan dard General Requirements for Liquid-Immersed Distribution, Power, and Regulating Transformers (ANSI).
3.
Power Transformer Maintenan ce and Testing, General Physics Corporat ion. 1990.
4.
Guidelines for th e Life Extension of Substa tions EPRI, TR-105070. April 1995.
5.
Transformer Maintenance Guide, by J .J Kelly, S.D. Myers, R.H. Parrish, S.D. Meyers Co. 1981.
6.
Transformer General Gasketing Procedures, by Alan Cote, S.D. Meyers Co. 1987.
7.
NFPA 70B-1998 Recommended Pra ctice for Electrical Equipment Maintenance.
8.
Bushing Field Test Guide, Doble Engineering Company. 1966.
9.
Testing and Maintenance of High-Voltage Bushings, FIST 3-2, Bureau of Reclamation. 1991.
10. IEEE Standa rd C57.19.00, 1991 General Requirements and Test Procedure for Outdoor P ower Appar at us Bush ings. 11.
IEEE Sta nda rd C57.104-1991 Guide for th e Interpr etat ion of Gases Genera ted in OilImmersed Transformers.
12.
Inter na tiona l Electr otechnical Comm ission (Draft IEC 60599 Edition 2), Mineral OilImpr egnated Electr ical E quipment in Service-Int erpret at ion of Dissolved an d F ree Gas Ana lysis. 1997.
13.
Dissolved Gas Analysis of Tran sform er Oil, by John C. Drotos, John W. Port er, Randy Stebbins, published by the S.D. Meyers Co. 1996.
14.
IEEE Sta nda rd C57.94, 1982, Recomm ended Pra ctice for Inst allation, Applicat ion, Operation an d Maint enan ce of Dry-Type Genera l Pu rpose Distribut ion a nd Power Transformers.
15.
Criteria for th e Inter preta tion of Data for Dissolved Gases in Oil from Tran sform ers (A Review), by Pa ul Gr iffin, Doble En gineer ing Co. 1996.
16.
Maintena nce of High Voltage Transform ers, by Mart in Heath Cote Associates, London, En gland. 1989.
17.
Therm al Monitors and Loading, by Ha rold Moore, from Tran sform er Perform ance Monitoring and Diagnostics EP RI. Sept ember 1997. 80