TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 16, No. 3, Dece
mbe
r
2
015, pp. 409
~ 416
DOI: 10.115
9
1
/telkomni
ka.
v
16i3.937
6
409
Re
cei
v
ed
Jul
y
13, 201
5; Revi
sed
No
ve
m
ber 15, 201
5; Acce
pted
No
vem
ber 2
7
,
2015
Power Transformer Incipient Faults Diagnosis Based
on Dissolved Gas Analysis
Osama E. Gouda*
1
, Saber M. Saleh
2
, Salah
Hamdy
EL-Hosh
y
3
1
Departme
n
t of Electrical Po
wer and Mac
h
in
es,
F
a
cult
y
of
Engi
neer
in
g, Cairo U
n
ivers
i
t
y
2
Departme
n
t of Electrical Po
wer and Mac
h
in
es, F
a
cult
y
of
Engi
neer
in
g, F
a
youm U
n
ivers
i
t
y
3
Middle E
g
ypt electricit
y zo
ne
, Eg
y
p
ti
an El
ec
tricit
y
T
r
ansmission C
o
mpa
n
y
,
El-F
a
y
o
u
m
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: Prof_ossama
11@
ya
ho
o.co
m
A
b
st
r
a
ct
Incipi
ent fa
ult
d
i
ag
nosis
of
a
p
o
w
e
r transfor
m
er is
gre
a
tly
inf
l
ue
nced
by
the
con
d
itio
n
asse
ssme
n
t
of its insulati
on
system oil a
n
d
/or pap
er ins
u
latio
n
.
Dissolv
e
d
gas-i
n-oi
l an
alysis (DGA) is
one of the mo
st
pow
erfull tec
h
niq
ues for the
detectio
n
of i
n
cipi
ent
fau
l
t cond
ition w
i
thi
n
oil-i
m
mers
ed
transformers. T
h
e
transformer da
ta has be
en
ana
ly
z
e
d
usin
g key gas
es, Doer
nen
bur
g, Rog
e
r, IEC and Duv
a
l tria
n
g
l
e
techni
qu
es. T
h
is pa
per
intro
d
u
ce
a MAT
L
A
B
pro
g
ra
m to
h
e
lp
in
un
ificati
o
n DGA i
n
terpr
e
tation tec
hni
qu
es
to investig
ate the accur
a
cy of these techn
i
q
ues in i
n
terpr
e
ting the transf
o
r
m
er co
nd
ition
and to prov
id
e th
e
best sug
gestio
n
for the type
of the fault w
i
thin th
e
transfo
rmer
base
d
on
fault perce
nta
ge. It propos
e
s
a
prop
er
ma
inten
ance
actio
n
b
a
s
ed o
n
DGA r
e
sults w
h
ic
h is
u
s
eful
for pl
ann
i
ng an appr
opri
a
te mai
n
ten
a
n
c
e
strategy to kee
p
the pow
er tra
n
sfo
rmer in
ac
ceptab
le co
nd
ition. T
he
ev
al
u
a
tion is c
a
rrie
d
out on DGA d
a
ta
obtai
ne
d from
352 o
il sa
mple
s has be
en su
mmar
i
z
e
d
i
n
to
46 sa
mpl
e
s th
at have b
e
e
n
collect
ed fro
m
a
3
8
different transf
o
rmers of differ
ent rating a
nd
different life sp
an.
Ke
y
w
ords
:
di
ssolve
d
gas a
nalysis (DGA),
fault diag
nos
i
s
, pow
er transformer, ins
u
l
a
ti
ng oi
l, conditi
on
assess
m
e
nt
Copy
right
©
2015 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
Powe
r tran
sforme
rs a
r
e
th
e mo
st exp
e
n
s
ive
a
n
d
imp
o
rtant
equi
p
m
ent in
ele
c
tric p
o
wer
system
s that serve
s
to convert
the p
o
we
r with different level
of voltage. During its lifetime,
transfo
rme
r
i
n
sul
a
tion is
exposed to mainly el
ect
r
ical, me
chan
ical an
d the
r
mal st
re
sse
s
.
Therefore, th
e life of a tra
n
sf
ormer
dep
end
s mainly
on the
con
d
ition of the in
sulation a
nd its
ability to withstand the
above stre
sses,
during its normal operation.
The m
a
in
reason behind t
he
stre
sse
s
can
be du
e to agi
ng of in
sulati
on an
d the m
o
st critical re
aso
n
is
a sho
r
t circuit eve
n
t.
As the health
of the oil implies the he
a
l
th of
the transformer, the
oil should b
e
sampl
ed a
nd
tested re
gula
r
ly to evaluate the oil cond
ition and
to determin
e
the possibl
e fault problem
s [1, 2].
Therefore, in
cipie
n
t fault detectio
n
of transfo
rm
e
r
s is ne
ce
ssary
in orde
r to avoid ele
c
trical
power sy
ste
m
fault as qui
ckly a
s
po
ssi
ble.
The tran
sformer i
n
sulatio
n
can
be
cla
ssifie
d
into
m
i
neral
oil
and
soli
d in
sul
a
tion [3, 4].
Any wea
k
ne
ss of in
sul
a
tion may re
su
lt in failure of the transf
o
rmer [5]. Ins
u
lating oil in a
transfo
rme
r
durin
g op
era
t
ion is expo
sed to
a
co
mbination
of heat, oxygen and el
ectri
c
a
l
discha
rge,
which
may lea
d
to its d
e
g
r
a
dation e
s
p
e
ci
ally throu
gh t
he p
r
o
c
e
ss
o
f
oxidation [6
-8].
Whe
n
an in
cipie
n
t fault occurs in po
wer tran
sfor
mer, either
electri
c
al o
r
thermal, cau
s
ing
decompo
sitio
n
of the tran
sformer
oil an
d a num
b
e
r
of gase
s
a
r
e
gene
rated
an
d dissolved i
n
to
the oil. Su
ch
these
ga
se
s
are
mainly in
cl
ud
e Hyd
r
o
g
en (H2
)
, met
hane
(CH4),
Ethane
(C2
H
4),
Ethylene (CH6), a
c
etylene
(C2H2), carbon m
onox
id
e (CO) and
carbo
n
dioxid
e
(CO2) and
the
non-fa
ult ga
ses
are
nitrog
en
(N2
)
, oxy
gen
(O2
)
. T
r
ansfo
rme
r
oil
ca
n d
e
comp
ose
CO a
nd
CO2
as a re
sult of
its oxidation [8].
Dissolve
d g
a
s
a
nalysi
s
(DGA) is on
e
of the
mo
st acce
ptable
techni
que
s
u
s
ed
for
detectin
g
an
d evaluatin
g
gasse
s di
sso
l
ved from mi
neral
oil an
d
also f
r
om
solid in
sulatio
n
of
internal
in
cipi
ent faults in
the p
o
wer tra
n
sformer
based on the oil
s
a
mples [9, 10]. It is
c
r
itic
ally
importa
nt to
monito
r the
inci
pient
of tran
sformer fault de
pen
ding
on th
e
type an
d t
h
e
con
c
e
n
tration
s
of these g
a
se
s form
ation. DGA int
e
rp
retation t
e
ch
niqu
es
such a
s
Key
gas
techni
que,
Doern
enb
ug t
e
ch
niqu
e, Roger ratio
te
chni
que, IEC ratio
s
te
ch
n
i
que, a
nd
Duval
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 16, No. 3, Dece
mb
er 201
5 : 409 – 416
410
triangle te
chn
i
que [8], [11-14] are different techni
q
u
e
s
for the dete
c
tion of incipi
ent transfo
rm
er
fault conditio
n
s.
In this pape
r,
the dissolve
d gasse
s were interp
reted
usin
g seve
ral
techniq
u
e
s
. Different
techni
que
s gi
ve different analysi
s
re
sult, and it is face
d with so mu
ch diverse inf
o
rmatio
n. These
techni
que
s of
interp
retatio
n
of the fault
gases
are inv
e
stigate
d
. Th
e study
wa
s
done to
evalu
a
te
the accu
ra
cy
of each te
chniqu
e in p
r
edictin
g the
fault based
on a fault p
e
rcentag
e u
s
ing
MATLAB program.
2.
Samples of
Ev
aluated Tr
ansformers
The
evaluati
on i
s
ca
rrie
d
out o
n
DGA
data
obtain
e
d
from
35
2 o
il sa
mple
s h
a
s
bee
n
summ
ari
z
ed
into 4
6
sam
p
les that
ha
ve bee
n
coll
ected
from
a
38
differe
nt
transfo
rme
r
s
of
different ratin
g
and differe
nt life span
s
as given in
T
able 1. It is difficult to determine
wheth
e
r
a
transfo
rme
r
i
s
beh
aving n
o
rmally if it has no p
r
evio
us di
ssolved
gas hi
sto
r
y. For an
alyzin
g
th
e
aging d
e
g
r
ee
of the transf
o
rme
r
for n
o
rmal state,
we
can
cla
ssify the no
rmal
co
ndition a
s
a f
our
status
con
d
i
t
ions such as: No
rmal
Operat
ion,
Caution
(investigate
)
, Abnorm
a
l (more
investigatio
n), and Da
nge
r
(nea
rin
g
failure). The
DG
A
guide to
cla
s
sify risks to tra
n
sformers
with
no previous
probl
em
s
has been developed by the IEEE
[5,
1
3
]. It uses combinations
of
individual g
a
ses a
nd total
combu
s
tible g
a
s
con
c
e
n
trat
ion. Thi
s
guid
e
is n
o
t unive
rsally a
c
cept
ed
and only on
e of the tools is us
ed to eval
uate tran
sformers.
Table 1. Tran
sform
e
r detail
s
Transf.
No. of tra
n
sform
e
r
Y
e
a
r
of Manuf
acture
Rating MVA
Voltages KV
A
7 – transform
er
3- transfo
rmer
(1
981)
3- transfo
rmer
(2
005)
1- transfo
rmer
(2
005)
75
125
75
220/66/11
B
2 – transform
er
2- transfo
rmer
(1
982)
75
220/132
C
23 – transfo
rmer
6- transfo
rmer
(1
990)
2- transfo
rmer
(1
997)
2- transfo
rmer
(2
000)
4- transfo
rmer
(2
004)
5- transfo
rmer
(2
005)
2- transfo
rmer
(2
008)
2- transfo
rmer
(2
011)
25 66/11
D
3 – transform
er
1- transfo
rmer
(1
978)
2- transfo
rmer
(1
980)
20 66/11
E
3 – transform
er
1- transfo
rmer
(1
978)
2- transfo
rmer
(1
983)
12.5 66/11
Table 2 [5], [
13] gives one set
of guideli
nes
based on good utility practi
ce that i
s
useful
for dete
r
mini
ng the overal
l health of a
power tr
a
n
sf
orme
r ba
se
d
on the total con
c
e
n
tration
of
comb
ustibl
e gases. The fo
ur co
ndition
s
are defin
ed b
e
low:
Table 2. Di
ssolved key con
c
entration lim
its in (ppm
)
G
a
ses in (
PPM)
G
a
ses Status
Normal
Caution
Abnormal
Danger
H
y
dr
ogen (PPM)
100
101-700
701-1800
>
1800
Methane (PPM
)
120
121-400
401- 1000
> 1000
Acety
l
en
e (PPM)
35
36-50
51-80
> 80
Eth
y
lene (PPM
)
50
51-100
101-200
> 200
Ethane (PPM)
65
66-100
101-150
> 150
Carbon M
ono
xide PPM)
350
351-570
571-1400
> 1400
Carbon Dioxide
(
PPM)
a
2500
2501-400
0
4001-000
0
>10000
TDC
G
(PPM)
720
721-1920
1921-463
0
>4630
a
it is not a combustibl
e ga
s
3.
Proposed
Accura
te DGA Technique
It is worth
to
mention
that
in
som
e
ca
ses, fo
r e
a
ch i
ndividual
trad
itional inte
rp
retation
techni
que
s a
ppro
a
ch su
ch as "key ga
s, rog
e
r'
s
rati
o, Doe
r
nen
b
u
rg ratio, IEC ratio a
nd
Duval
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Powe
r Tra
n
sf
orm
e
r Incipi
e
n
t Faults Dia
gno
sis Ba
sed
on Dissol
ve
d
Gas…
(O
sa
m
a
E. Gouda
)
411
triangle"
hav
e bee
n give
n
different fa
ult conditio
n
s for the
same
sampl
e
u
n
it. To avoid thi
s
,
the
sug
g
e
s
ted te
chni
que
give
s the
app
rop
r
iate fault ba
sed on fa
ult p
e
rcentag
e wh
ich i
s
given
b
y
all
the above exi
s
ting DGA interp
retation te
chni
que
s with
out any overl
appin
g
.
Firstly, che
c
k the
limit value
of the
comb
ustibl
e
and
non
co
mbustibl
e g
a
s
e
s
a
r
e
appli
c
able
o
r
not ba
se
d
on
the
qua
ntity of ea
ch
ga
s
concentratio
n
s in
ppm
(p
art
s
p
e
r million
)
as
an input data
for each diag
nosti
c tech
niq
ue as give
n in table 3qu
oted from [5, 13].
Table 3. Limit
con
c
entratio
n
s of dissolve
d gas
Ke
y
gas
Concentrations L
1
"ppm"
H
y
dr
ogen (H2)
100
Methane (
CH4)
120
Carbon m
ono
xide (C
O)
350
Acety
l
en
e (C2
H
2
)
1
Eth
y
lene (
C
2H4
)
50
Ethane (C2
H
6)
65
Table 4. Sug
geste
d DGA f
ault types co
des for
accu
rate DGA tech
nique
Tec
hni
que
F1
F2
F3
F4
F5
F6
F7
Duval Normal
Thermal F
ault
<300 ºC
Thermal F
ault
300-700 ºC
Thermal
Fault >700
ºC
Lo
w
ener
g
y
discharge
and
High energ
y
discharge
PD
With mix
thermal and
electrical fault
Out
of
code
fault
Doernen
bug
Normal
Out of code
fault
Thermal decomp
o
sition
Arcing
Par
t
ial
Discharges
Out
of
code
fault
IEC Normal
Thermal fault of
low
tem
p
<150
ºC
Thermal fault of
low
tem
p
bet
w
een 15
0-
300 ºC
Thermal fault of
medium temp.
bet
w
een 30
0-
700 ºC
Thermal
fault of high
temp >700
ºC
Discharge
wi
t
h
l
o
w
energ
y
Sparking
Discharge
w
i
th high
energ
y
,
Arcing
Par
t
ial
discharge
wi
t
h
l
o
w
energ
y
density
Out
of
code
fault
Rogers
Normal
Thermal fault of
low
temperatu
r
e
range <150 ºC
Thermal fault of
temperatu
r
e
range 150
-200
ºC
Thermal fault of
temperatu
r
e
range 200
-300
ºC
Winding
circulating
current
Core/tank
circulating
current.
Insulated
conductor
overheat
Arc w
i
th
po
w
e
r
follow
s
through.
Sparking.
Flashover.
Par
t
ial
discharge
Par
t
ial
discharge
w
i
th t
r
acking
Out
of
code
fault
Ke
y
Gas
Normal
Out of code
fault
Overheating of
oil
Hot Spot
Arcing fault
Par
t
ial
Discharge
(Coro
na)
Out
of
code
fault
Accurate
Technique
Normal
Thermal F
ault
<300 ºC
Thermal F
ault
300-700 ºC
Thermal
Fault >700
ºC
Arcing fault
Par
t
ial
discharge
Out
of
code
fault
Secon
d
ly, the same
sam
p
ling oil unit
is analyzed
by each co
mputeri
z
e
d
individual
techni
que to
determi
ne th
e possibl
e fault types. As
an example, f
o
r the same
sampl
e
unit , Ke
y
Gas te
ch
niqu
e got dia
gno
sis
co
ndition
as "Aci
ng in
Oil", Roge
rs
Ratio te
chni
q
ue got di
agn
osi
s
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 16, No. 3, Dece
mb
er 201
5 : 409 – 416
412
con
d
ition a
s
"Arc
with Powe
r Foll
ow Thro
ugh", I
E
C Ratio te
chni
que
got
con
d
ition a
s
"No
Predi
ction", a
nd Do
ernen
b
u
rg
Ratio te
chniqu
e got
di
agno
si
s co
nd
ition as "Arci
ng" and
Duv
a
l
Trian
g
le te
ch
nique
got dia
gno
sis
co
ndit
i
on a
s
"D
i
s
ch
arge
of Hig
h
Energy". Th
erefo
r
e, for t
h
e
fault identification with re
spe
c
t to ea
ch tradition
al techni
que, th
e faults a
r
e
categ
o
ri
zed i
n
to
seven type
s
of faults whi
c
h is a
ssi
gned
with
fault co
des F
1
-F
7 a
s
su
gge
sted
in Table 4. T
he
next step is t
o
com
pare al
l incipie
n
t fault types
using
each te
chni
que. The
n
, the con
c
lu
sio
n
is
given for final
fault type of t
he te
sted tra
n
sforme
r oil
same
sampl
e
unit ba
sed
on
the pe
rcenta
ge
given by each techinq
ue. Finally, using
MATLAB so
ftware all dia
gno
stic re
sult
s of individu
al
techni
que a
n
d
Accu
rate DGA techniq
u
e
result a
r
e di
splayed.
Figure 1. The
flowch
art of the pro
c
e
d
u
r
e
of Accurate DGA techniq
ue
The flowch
art
on a PC
usi
n
g MATLAB software
given
in Figu
re 1 i
s
use
d
in thi
s
pape
r to
rep
r
e
s
ent the
results of all
mentioned t
e
ch
niqu
es. T
h
e Accurate DGA techniq
ue pro
g
ram will
focu
s o
n
the
fault examina
t
ion an
alysisi
dentificat
io
n
usin
g fault p
e
r
ce
ntage
an
a
l
ysis to
find t
he
accurate fault
result, in ord
e
r to diag
no
se the
incipi
en
t faults of the suspe
c
ted transfo
rme
r
s
a
n
d
sug
g
e
s
t prop
er mainte
nan
ce a
c
tion
s as
soo
n
as p
o
ssi
b
le.
4. Resul
t
s
and
Analy
s
is
Table 5
sho
w
s the diag
no
sis re
sult
s accordin
g to hea
lthy condition
s
and a
c
cu
rat
e
DGA
techni
que. Th
e develop
ed
accurate DG
A techniq
ue
(whi
ch is
deri
v
ed from indi
vidual traditio
n
al
techni
que
s) e
ffectively con
c
lud
e
the
ap
prop
riat
e
faul
t of the tradit
i
onal i
n
terp
re
tation ap
pro
a
c
h
based on fa
ult percenta
g
e
without a
n
y
overlappin
g
as
it is cl
ea
r in Tabl
e 5. The su
gge
sted
techni
que
is
based
on th
e
fault pe
rcent
age th
at is
m
a
inlydetermin
ed by
usi
n
g t
he exi
s
ting
DGA
interp
retation
techni
que
s
whi
c
h a
r
e li
st
ed in T
able
4
without any overlap
p
ing,
f
o
r that
re
sion
it is
more
a
c
curat
e
than
any o
ne of the
m
and it i
s
mo
re helpful
to
detect th
e transfo
rme
r
fa
ults
comp
ared
wi
th tech
nique
s
sug
gest
e
d
by Duval,
M. [8], Key Gas, IEC ratio, Doe
r
n
enb
urg
ratioan
d R. R. Rogers [14].
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Powe
r Tra
n
sf
orm
e
r Incipi
e
n
t Faults Dia
gno
sis Ba
sed
on Dissol
ve
d
Gas…
(O
sa
m
a
E. Gouda
)
413
Table 5. Re
sult of fault types code
s for i
ndivi
dual an
d
accurate
DG
A techniqu
e a
nd tran
sform
e
r
st
at
e
Transf
ormer
State
Accur
ate
Tec
h
n
ique
Traditional Tech
niques
G
a
ses in (
ppm)
n
o
Du
val
IE
C
Rog
er's
Doerne
nburg
K
ey
G
as
CO2
٭
CO
C
2
H
6
C
2
H
4
C
2
H
2
CH
4
H
2
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F3
F1
F
7
1453
130
22
58
0
65
41
1.
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Normal
F1
F1
F
3
F7
F1
F
7
1362
268
9
17
0
66
17
2.
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F1
F
7
533
143
36
186
0
48
20
3.
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
1050
496
73
579
0
306
179
4.
Norm
al
Cauti
on
Cauti
on
Dang
er
Norm
al
Cauti
on
Cauti
on
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F3
F1
F
7
5398
1767
37
52
0
177
73
5.
Abno
rmal
Dang
er
Norm
al
Cauti
on
Norm
al
Cauti
on
Norm
al
Arcing
fault
F5
F5
F
5
F7
F1
F
7
1070
167
12
102
187
43
210
6.
Norm
al
Norm
al
Norm
al
Abno
rmal
Dang
er
Norm
al
Cauti
on
Therm
al Fault
>700
ºC
F4
F4
F
7
F7
F1
F
4
2692
1469
12
65
0
7
39
7.
Cauti
on
Dang
er
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F3
F1
F
7
926
132
19
57
0
96
37
8.
Norm
al
Norm
al
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F7
F1
F
7
2678
184
27
150
0
147
46
9.
Cauti
on
Norm
al
Norm
al
Abno
rmal
Norm
al
Cauti
on
Norm
al
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F7
F1
F
7
2633
390
20
51
0
63
18
10.
Cauti
on
Cauti
on
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Arcing
fault
F5
F5
F
5
F7
F1
F
7
308
138
9
60
98
30
200
11.
Norm
al
Norm
al
Norm
al
Cauti
on
Dang
er
Norm
al
Cauti
on
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F1
F
7
1206
538
16
76
0
62
28
12.
Norm
al
Cauti
on
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
3
F7
F3,F4
F
7
1720
550
103
136
0
78
68
13.
Norm
al
Cauti
on
Abno
rmal
Abno
rmal
Norm
al
Norm
al
Norm
al
Therm
al Fault
300-
700 ºC
F3
F3
F
4
F3
F3,F4
F
7
4497
1358
23
101
0
143
54
14.
Abno
rmal
Abno
rmal
Norm
al
Abno
rmal
Norm
al
Cauti
on
Norm
al
Therm
al Fault
300-
700 ºC
F3
F4
F
3
F3
F1
F
7
1114
229
27
62
0
47
19
15.
Norm
al
Norm
al
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Arcing
fault
F5
F5
F
5
F7
F5
F
7
1909
768
31
89
237
70
678
16.
Norm
al
Abno
rmal
Norm
al
Cauti
on
Dang
er
Norm
al
Cauti
on
Therm
al Fault
>700
F4
F4
F
4
F7
F1
F
7
2574
323
63
338
0
277
86
17.
Cauti
on
Norm
al
Norm
al
Dang
er
Norm
al
Cauti
on
Norm
al
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 16, No. 3, Dece
mb
er 201
5 : 409 – 416
414
ºC
Therm
al Fault
>700
ºC
F4
F4
F
2
F7
F1
F
4
1854
571
45
77
0
19
20
18.
Norm
al
Abno
rmal
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
300-
700 ºC
F3
F4
F
3
F3
F1
F
7
1814
424
42
58
0
49
25
19.
Norm
al
Cauti
on
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F1
F
7
987
312
64
379
0
150
53
20.
Norm
al
Norm
al
Norm
al
Dang
er
Norm
al
Cauti
on
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
5675
1309
80
574
0
241
114
21.
Abno
rmal
Abno
rmal
Cauti
on
Dang
er
Norm
al
Cauti
on
Cauti
on
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
1625
304
73
386
0
288
99
22.
Norm
al
Norm
al
Cauti
on
Dang
er
Norm
al
Cauti
on
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
2819
783
628
5376
298
2648
1374
23.
Cauti
on
Abno
rmal
Dang
er
Dang
er
Dang
er
Dang
er
Abno
rmal
Therm
al Fault
>700
ºC
F4
F4
F
3
F7
F3,F4
F
7
3197
645
100
136
0
83
34
24.
Cauti
on
Abno
rmal
Cauti
on
Abno
rmal
Norm
al
Norm
al
Norm
al
Therm
al Fault
<300
ºC
F2
F3
F
2
F2
F1
F
7
2576
712
61
17
0
132
20
25.
Cauti
on
Abno
rmal
Norm
al
Norm
al
Norm
al
Cauti
on
Norm
al
Therm
al Fault
<300
ºC
F2
F3
F
2
F2
F1
F
7
5197
1046
83
16
0
138
13
26.
Abno
rmal
Abno
rmal
Cauti
on
Norm
al
Norm
al
Cauti
on
Norm
al
Arcing
fault
F5
F5
F
5
F7
F5
F
7
5346
459
38
54
126
93
762
27.
Abno
rmal
Cauti
on
Norm
al
Cauti
on
Dang
er
Norm
al
Abno
rmal
Therm
al Fault
300-
700 ºC
F3
F4
F
3
F3
F3,F4
F
7
2326
718
65
139
0
116
43
28.
Norm
al
Abno
rmal
Cauti
on
Abno
rmal
Norm
al
Norm
al
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
1050
496
73
579
0
306
179
29.
Norm
al
Cauti
on
Cauti
on
Dang
er
Norm
al
Cauti
on
Cauti
on
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F3
F1
F
7
5484
1405
38
51
0
141
57
30.
Abno
rmal
Dang
er
Norm
al
Cauti
on
Norm
al
Cauti
on
Norm
al
Normal
F1
F1
F
1
F7
F1
F
4
3929
2723
34
15
0
8
40
31.
Cauti
on
Dang
er
Norm
al
Norm
al
Norm
al
Norm
al
Norm
al
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F7
F3,F4
F
7
1103
4
1475
121
222
0
283
35
32.
Dang
er
Dang
er
Abno
rmal
Dang
er
Norm
al
Cauti
on
Norm
al
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F7
F1
F
7
1126
3
1769
29
87
0
159
15
33.
Dang
er
Dang
er
Norm
al
Cauti
on
Norm
al
Cauti
on
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
1129
8
1309
114
493
0
159
55
34.
Dang
er
Abno
rmal
Abno
rmal
Dang
er
Norm
al
Cauti
on
Norm
al
Therm
al Fault
<300
ºC
F2
F3
F
2
F2
F1
F
7
3896
869
67
52
0
123
37
35.
Cauti
on
Abno
rmal
Cauti
on
Cauti
on
Norm
al
Cauti
on
Norm
al
Arcing
F5
F5
F
F7
F5
F
5346
459
110
293
288
191
723
36.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Powe
r Tra
n
sf
orm
e
r Incipi
e
n
t Faults Dia
gno
sis Ba
sed
on Dissol
ve
d
Gas…
(O
sa
m
a
E. Gouda
)
415
fault
5
7
Abno
rmal
Cauti
on
Abno
rmal
Dang
er
Dang
er
Cauti
on
Abno
rmal
Therm
al Fault
<300
ºC
F2
F4
F
2
F2
F1
F
7
4647
687
78
58
0
15
7
37.
Abno
rmal
Abno
rmal
Cauti
on
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F1
F
7
3566
803
12
54
0
51
30
38.
Cauti
on
Abno
rmal
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
300-
700 ºC
F3
F4
F
3
F3
F1
F
7
5175
771
33
77
0
56
31
39.
Abno
rmal
Abno
rmal
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F3
F3,F4
F
7
5240
1150
68
192
0
226
109
40.
Abno
rmal
Abno
rmal
Cauti
on
Abno
rmal
Norm
al
Cauti
on
Cauti
on
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
3085
897
66
505
0
279
137
41.
Cauti
on
Abno
rmal
Cauti
on
Dang
er
Norm
al
Cauti
on
Cauti
on
Therm
al Fault
300-
700 ºC
F3
F3
F
3
F3
F1
F
7
4738
735
36
70
0
119
59
42.
Abno
rmal
Abno
rmal
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
Therm
al Fault
300-
700 ºC
F3
F3
F
4
F3
F3,F4
F
7
5418
1076
68
232
0
242
151
43.
Abno
rmal
Abno
rmal
Cauti
on
Dang
er
Norm
al
Cauti
on
Cauti
on
Par
t
ial
dischar
ge
F6
F5
F
7
F6
F6
F
7
5346
459
73
54
14
77
870
44.
Abno
rmal
Cauti
on
Cauti
on
Cauti
on
Abno
rmal
Norm
al
Abno
rmal
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F3,F4
F
7
2160
674
146
1092
0
575
376
45.
Norm
al
Abno
rmal
Abno
rmal
Dang
er
Norm
al
Abno
rmal
Cauti
on
Therm
al Fault
>700
ºC
F4
F4
F
4
F3
F1
F
7
823
78
11
86
0
40
17
46.
Norm
al
Norm
al
Norm
al
Cauti
on
Norm
al
Norm
al
Norm
al
In so
me
ca
ses, A
c
curate
DGA T
e
ch
nique
may l
e
ad to
more t
han
one
result (e.g.
Accu
rate
T
e
ch
niqu
e
in sampl
e
s nu
mber 2,
6,
7
,
9, 10, 11, 33, 44). T
o
avoid this i
n
th
e
developed proposed tec
hnique, the output will
be sel
e
cted based on
the worst
case.
Con
s
e
quently
, based on
MATLAB Co
de the abov
e analysi
s
is summa
rized
in table 5.
The
internal
in
cipi
ent faults
dia
gno
sis
of po
wer tran
sfo
r
mer i
s
al
so
a
c
compli
she
d
by evaluation
of
transfo
rme
r
condition u
s
in
g individual
and T
D
CG
concentratio
n
s for determi
n
i
ng the overall
health
of a
power t
r
an
sf
orme
r b
a
sed
on fo
ur stat
us co
ndition
s:
No
rmal Op
eration, Ca
ution,
Abnorm
a
l, an
d Dange
r
(n
earin
g failu
re
). So, the
re
su
lt of thi
s
work is u
s
eful
for pla
nnin
g
an
approp
riate
maintena
nce
strategy to
kee
p
th
e
po
wer tra
n
sfo
r
mer i
n
a
c
ce
ptable
co
ndit
i
ons.
Therefore, in
cipie
n
t fault detectio
n
of transfo
rm
e
r
s is ne
ce
ssary
in orde
r to avoid ele
c
trical
power sy
ste
m
fault as qui
ckly a
s
po
ssi
ble.
5. Conclu
sion
This
pap
er i
n
trodu
ce
s a
n
accurate
DG
A techni
que
for id
entifying
the inte
rnal
i
n
cipi
ent
faults in
the
p
o
we
r tran
sformer
ba
sed
on
the fault
pe
rcentage
of di
ssolved
ga
sse
s
relea
s
e
d
fro
m
transfo
rme
r
oil and
soli
d insul
a
tion
using
pra
c
t
i
cal oil
sam
p
les
colle
cte
d
from diffe
rent
transfo
rme
r
s. The te
ch
niqu
es
of interpre
tation of the f
ault ga
se
s a
r
e inve
stigate
d
an
d comp
a
r
ed
with ea
ch
other b
a
sed o
n
the Accu
rate DGA te
chni
que
whi
c
h is d
e
rived
from individ
ual
techni
que
s t
o
evaluate t
he a
c
cura
cy
of each
techniqu
e in p
r
edictin
g the
fault based
on a
MATLAB pro
g
ram a
s
sho
w
n in Ta
ble 5
.
Furtherm
o
re
, the deteriorated tran
sformer by mea
n
s of
this analy
s
is can b
e
carefully focu
se
d for t
he ap
prop
riate m
a
i
n
tenan
ce a
s
con
d
ition-ba
sed
maintena
nce before
seve
re
damage o
ccurs.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 16, No. 3, Dece
mb
er 201
5 : 409 – 416
416
Referen
ces
[1]
A Abu-Sia
da, S Islam. A ne
w
appro
a
ch to id
ent
if
y
p
o
w
e
r transformer criti
c
alit
y
an
d asse
t managem
ent
decisi
o
n
bas
e
d
on
diss
olv
e
d gas-
i
n-
oil
a
nal
ysis.
IEEE
Transactions
on Dielectrics
and Electric
al
Insulation
. 20
1
2
; 19: 100
7-10
12.
[2]
Ena Nar
ang,
Er Shivanis
e
h
gal. F
ault Det
e
ction T
e
chniq
ues for Mainte
nanc
e Usin
g Dissolv
ed g
a
s
Anal
ys
is.
Internatio
nal J
ourn
a
l of Engi
ne
er
i
ng Res
earch &
T
e
chnol
ogy (IJERT
)
. 2012; 1
(
6): 1-7.
[3]
Sherif SM Gh
o
neim, Sa
ye
d A
W
a
rd. Disso
lv
ed g
a
s An
al
ysi
s
an Ear
l
y Ide
n
tificatio
n
of
T
r
ansformer
Faults.
Advan
c
es in Electric
al Engi
ne
erin
g
Systems (AEE
S).
2012; 1(3).
[4]
Andri F
e
bri
y
a
n
to, T
apan Ku
mar Sah
a
. Oil-
immersed P
o
w
e
r T
r
ansformers Con
d
itio
n Di
agn
osis
w
i
t
h
Limite
d Disso
lved Gas Anal
ysis (DGA
) Data.
Australasi
an Un
iv
ersities Pow
e
r
Engin
eeri
n
g
Confer
ence (A
UPEC)
. 200
8: 73.
[5]
IEEE Guide for
the Interpreta
ti
on of Gases G
ener
ated i
n
Oil
-
Immersed T
r
ansformers. C5
7.104-
20
08..
Institute of Electrical an
d Elec
tronics Eng
i
ne
ers
, Inc. Ne
w
Y
o
rk: 2008; 9-
27
.
[6]
Siva Sarma,
GNS Kaly
ani.
ANN Appr
oac
h for Con
d
itio
n Monitor
i
ng
o
f
Pow
e
r T
r
ansformers
usin
g
DGA
. IEEE Regio
n
10 C
onfer
ence,
T
E
NCON 200
4. 200
4: 444-
447.
[7]
M Duval. Ne
w
techni
qu
es for dissolv
ed g
a
s-i
n
-oil a
n
a
l
ysis.
IEEE Electrical
Insulation Maga
z
i
ne.
200
3;
19: 6-15.
[8]
Duval M. A Revie
w
of Fa
u
l
ts Detectab
le
b
y
Gas-in-Oi
l
Anal
ys
is in T
r
ansformers.
IE
EE Electrical
In
su
l
a
tio
n
Maga
z
i
ne
. 200
2; 18(3).
[9]
DVSS Sivas
a
r
m
a, GNS Kalyani. ANN appr
oach f
o
r Co
ndition M
onit
o
ring
of Po
w
e
r
T
r
ansformers using
DGA.
IEEE Electrical Insulation Maga
z
i
ne
. 2
002: 12-
25.
[10] X Li
u, F
Z
hou,
F
Huang.
Res
earch o
n
on-l
i
n
e
DGA using F
T
I
R [po
w
er tra
n
sformer insu
l
a
tion testin
g].
200
2; 3: 1875-
188
0.
[11]
T
O
Rouse. Min
e
ral i
n
sul
a
tin
g
oil in tra
n
sform
e
rs
. IEEE Electr. Insul. Mag
. 1998; 14(
3): 6-1
6
.
[12] IEEE and IEC Cod
e
s to Interpret Incipi
ent Faults
in T
r
ansformers, Using
Gas in Oil Ana
l
y
s
is, b
y
R.R.
Rog
e
rs C.E.G.B,
T
r
ansmissio
n
Divisi
on, Guil
ford, Engla
nd. 199
5.
[13]
IEEE Po
w
e
r E
ngi
neer
in
g So
ciet
y
.
C57.
104
-199
1.
IEEE
Guide
for the In
terpr
e
tation of Gas
e
s
Generate
d
in Oil-Immers
ed T
r
ansformers.
1
9
92
[14]
RR Ro
gers. I
EEE and IEC
Cod
e
s to Inter
p
ret Inci
p
i
ent
Faults in T
r
an
sformers, Usin
g Gas in Oi
l
Anal
ys
is.
IEEE Transactions
on Electric
al Insulation
. 19
78; 13: 349-
35
4.
Evaluation Warning : The document was created with Spire.PDF for Python.