Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
6, N
o
. 4
,
A
ugu
st
2016
, pp
. 14
81
~
1
488
I
S
SN
: 208
8-8
7
0
8
,
D
O
I
:
10.115
91
/ij
ece.v6
i
4.1
107
2
1
481
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Comparative Study on the AC
Breakdown
Vol
t
age of Palm
Fatty Acid Ester Insulation
Oils Mixed With Iron Oxide
Nanoparticles
Mohd
Safwan
Moh
a
m
a
d,
H
i
dayat Z
a
inud
din,
S
h
arin
Ab Gh
ani,
Imr
a
n Su
tan
Ch
airul
High Voltag
e
En
gineer
ing Res
ear
ch Laborator
y,
F
acul
t
y
of
El
ectr
i
cal
Engin
eer
ing,
Universiti Tekn
ikal Malay
s
ia
Melaka, Hang
Tu
ah
Jay
a
, 76100
Durian
Tunggal, M
e
lak
a
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Apr 28, 2016
Rev
i
sed
Jun
30,
201
6
Accepte
d
J
u
l 14, 2016
Nowaday
s
, n
a
no
techno
log
y
h
a
s become one of the most important resear
ch
fields in both th
e acad
em
ia and industr
y
and i
t
has been shown
in previous
s
t
udies
that na
nos
cale m
a
t
e
ria
l
s
are benef
i
ci
a
l
for trans
f
orm
e
rs
. In this
regard,
the ob
je
ctiv
e of this stu
d
y
is
to com
p
ar
e the AC br
eakd
o
wn voltag
e
of palm fatty
acid es
te
r (PFAE) oils m
i
xe
d with iron oxide (Fe
3
O
4
)
nanoparticles.
The PFAE-based nanof
luids ar
e prepar
ed bas
e
d on two
methods suggested in
liter
a
ture.
Method I: the con
centr
atio
n of Fe
3
O
4
nanopart
icl
e
s is
weight-b
ased (
g
/l), i
.
e. gram
f
o
r each
litr
e of
PFAE oil
.
Method II: the
concen
tration of
Fe
3
O
4
nanoparticles is based on volume-
fraction (%). The AC breakdo
wn volta
ge test is conducted on
the PFAE-
based nanoflu
ids in accordan
ce
with th
e ASTM
D1816 standard
test method
.
Weibull statistical an
aly
s
is is
carried
out to
analy
s
e th
e AC
breakdown
voltag
e
of fr
esh
PFAE oil and P
F
AE-based
nano
fluids. I
t
is foun
d that th
ere
is enhancement
of the AC break
down
voltage fo
r all PFAE-based nanofluids
with the exception of one samp
le prepar
ed usin
g Method II (0.01% Fe
3
O
4
nanoparticles).
Keyword:
B
r
eak
do
w
n
Vo
l
t
a
ge
Nanoflu
id
s
Pal
m
Fatty Acid
Ester
Transform
e
r Oil
Weibu
ll Statistical An
alysis
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
M
o
h
d
Sa
f
w
an
M
oham
a
d,
High Voltage
Engineeri
n
g
R
e
search
La
bora
tory,
Faculty of Elect
ri
cal
E
n
gi
nee
r
i
n
g,
Un
i
v
ersiti Tekn
ik
al Malaysia Melak
a
,
H
a
ng
Tu
ah
Jay
a
, 761
00
Du
r
i
an
Tun
g
g
a
l, Melak
a
, Malaysia.
Em
a
il: safwan_
m
o
h
d
9
0
@yah
oo
.co
m
1.
INTRODUCTION
Hi
g
h
V
o
l
t
a
ge
engi
neeri
ng i
s
wi
del
y
use
d
i
n
p
o
we
r sy
st
em
s, especi
al
l
y
for p
o
w
er
gene
rat
i
o
n
,
tran
sm
issio
n
an
d d
i
stri
b
u
tion. In
su
latio
n plays a p
r
o
m
in
en
t ro
le in
h
i
gh
v
o
ltag
e
syste
m
s su
ch as ro
tating
machines, power capacitors, powe
r tran
s
f
o
r
m
e
rs, switchg
e
a
rs, circuit b
r
e
a
kers a
nd
po
w
e
r cables to en
sure
t
h
at
t
h
e sy
st
em
s are i
n
best
p
e
rf
orm
a
nce. The l
i
f
espa
n o
f
hi
g
h
v
o
l
t
a
ge s
y
st
em
s
i
s
heav
i
l
y
depend
ent
on t
h
e
reliab
ility o
f
t
h
e insu
lation
of th
e system
s
an
d th
is i
n
cludes th
e i
n
su
latio
n m
a
terials [1
]-[3
]. Th
e i
n
su
latio
n
m
e
di
a used i
n
hi
g
h
v
o
l
t
a
ge s
y
st
em
s can be
gene
ral
l
y
cl
assi
fi
ed as s
o
l
i
d
s,
l
i
qui
ds a
n
d ga
ses. I
n
gene
ral
,
sol
i
d
and l
i
q
ui
d i
n
s
u
l
a
t
i
ng m
e
di
a are com
m
onl
y
used
f
o
r t
r
a
n
sfo
r
m
e
rs. Sol
i
d i
n
s
u
l
a
t
i
ng m
e
di
a i
n
cl
u
d
e
b
a
rri
ers
,
bl
oc
ks a
n
d
s
p
a
cers m
a
de of
c
e
l
l
u
l
o
se
press
b
oar
d
or
w
o
od
whe
r
eas l
i
q
ui
d
i
n
sul
a
t
i
n
g m
e
di
a i
n
cl
ude
pet
r
ol
eum
-
deri
ved m
i
nera
l
oi
l
s
, sy
nt
het
i
c l
i
qui
ds a
s
we
l
l
as veget
a
bl
e-
b
a
sed
o
ils
[4
]. Min
e
ral o
ils
are typ
i
cally u
s
ed
as
in
su
lation
liq
u
i
d
s
fo
r t
r
an
sfo
r
mers d
u
e
t
o
their ex
cellen
t
di
electric and c
o
oling
propertie
s, as well as low cost.
Mo
reo
v
e
r, m
i
n
e
ral o
ils are co
mmercially a
v
ailab
l
e in
th
e
m
a
rket
. H
o
we
ver
,
des
p
i
t
e
t
h
e bene
fi
t
s
o
f
m
i
neral
oi
l
s
, t
h
ese oi
l
s
are deri
ve
d
fr
om
pet
r
ol
eu
m
,
whi
c
h i
s
a no
n-
rene
wa
bl
e and n
o
n
-
s
u
s
t
ai
nabl
e so
urc
e
, and
t
h
eref
o
r
e, t
h
es
e oi
l
s
are not
a
l
ong t
e
rm
opt
i
on
fo
r i
n
s
u
l
a
t
i
ng t
r
a
n
s
f
o
r
m
e
rs [5]
–
[7]
.
F
o
r t
h
i
s
reas
on
, t
h
e
r
e i
s
a
critical n
eed
to d
e
v
e
l
o
p
alternativ
e in
su
lating
m
e
d
i
a in
rep
l
ace
m
e
n
t
o
f
con
v
e
n
tion
a
l min
e
ral o
ils. Pal
m
o
il is
d
e
ri
v
e
d fro
m
oil p
a
l
m
fru
its an
d it h
a
s great
p
o
t
en
tial fo
r
u
s
e as an
in
su
latin
g m
e
d
i
u
m
fo
r tran
sfo
r
m
e
rs.
Th
is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 4
,
Au
gu
st 2
016
:
14
81
–
1
488
1
482
o
il o
ffers a n
u
m
b
er o
f
b
e
n
e
fits sin
ce it
is b
i
o
d
e
grad
ab
le and
it h
a
s h
i
g
h
co
o
ling
stab
ility an
d
go
od
ox
i
d
atio
n
stab
ility [8
],[9
]. More im
p
o
r
t
a
n
tly, it is po
ten
tial sub
s
titu
te of m
i
n
e
ral o
ils b
ecau
s
e of its ex
cellen
t
d
i
el
ectric
p
r
op
erties. Pal
m
fatty ac
id
s ester (PFAE) was d
e
v
e
lop
e
d in
2
006
b
y
Lio
n
C
o
rp
oration
as in
su
latio
n o
il for
tran
sform
e
r [10
]
. PFAE is
no
t on
ly env
i
ronmen
tally fri
end
l
y
but
i
t
al
so
h
a
s su
pe
ri
or
i
n
s
u
l
a
t
i
on
pe
rf
or
m
a
nce.
M
u
ch e
f
f
o
rt
ha
s bee
n
m
a
de t
o
im
pro
v
e t
h
e
di
el
ect
ri
c per
f
o
r
m
a
nce o
f
i
n
s
u
l
a
t
i
on
oi
l
s
an
d
one
o
f
t
h
e
s
e
i
n
v
o
l
v
es
bl
e
n
d
i
ng t
h
e i
n
s
u
l
a
t
i
o
n
oi
l
s
wi
t
h
n
a
no
pa
rt
i
c
l
e
s. S
u
ch
m
i
xt
ures a
r
e k
n
o
w
n
a
s
n
a
no
fl
ui
ds
[1
1]
–[
1
6
]
.
Ev
en
tho
ugh
th
e con
c
ep
t o
f
n
a
noflu
id
s
was in
itiated
b
y
Ch
o
i
in
19
93
for th
erm
a
l co
n
ductiv
ity en
h
a
n
c
emen
t,
st
udi
es
o
n
t
h
e
di
el
ect
ri
c pr
o
p
e
rt
i
e
s o
f
na
n
o
f
l
ui
ds as
i
n
s
u
l
a
t
i
ng m
e
di
a ha
d
onl
y
beg
u
n
i
n
19
9
8
by
Se
gal
et
al
.
[
1
7
]
.
A
ccor
d
i
n
g
to Lv et al.
[1
3
]
, t
h
e
n
a
nopar
ticles is d
i
v
i
d
e
d in
to thr
ee
g
r
ou
p
s
i.e co
nd
u
c
tiv
e n
a
no
p
a
r
ticles,
sem
i
-con
duct
i
ve na
n
opa
rt
i
c
l
e
s an
d i
n
s
u
l
a
t
i
ng
nan
o
p
art
i
c
les. Th
ere is
a lack
of in
t
e
rest in
co
ndu
ctiv
e
nan
o
p
art
i
c
l
e
s am
ong researc
h
ers beca
use of
t
h
ei
r hi
g
h
t
e
nd
ency
t
o
aggl
o
m
erat
e. St
ati
s
ti
cal
t
echni
ques
can be
u
s
ed
to
d
e
termin
e th
e p
e
rforman
ce o
f
sam
p
le b
y
using
AC b
r
eak
down
v
o
ltag
e
test.
In th
is reg
a
rd
, the
m
a
in
purpose
of sta
tistical techniques is to estimate th
e AC
brea
kdown voltage (also known as the wi
thstand
v
o
ltag
e
)
b
a
sed on
th
e prob
abilit
y o
f
d
i
stributio
n
d
a
ta [18
]
.
Th
e
o
b
j
ectiv
e
o
f
t
h
is stud
y is to
co
m
p
are the AC
brea
kdown vol
tage
of PFAE-base
d nanofl
uids pre
p
a
r
ed
usi
n
g
two diffe
re
nt
m
e
thods t
h
a
t
will be
discus
sed i
n
t
h
e fol
l
o
wi
n
g
sect
i
on. T
h
e
PFAE
-
base
d n
a
no
fl
ui
ds are
pre
p
are
d
by
d
i
spersi
ng i
r
on
oxi
de
na
no
pa
rt
i
c
l
e
s
(Fe
3
O
4
), a type o
f
co
ndu
ctive n
a
nop
articles in
PFAE o
ils.
W
e
ibu
ll statis
tical an
alysis i
s
th
en
carried
o
u
t
t
o
analyse and i
n
terpret the
data.
2.
EX
PER
I
M
E
NT PR
EPA
RATION
2.
1.
Sample
Preparati
o
n
The
nat
u
ral
est
e
r use
d
fo
r t
h
i
s
st
udy
was
pa
lm
fat
t
y
aci
d est
e
r (P
FAE
)
,
whi
c
h i
s
a
pr
o
duct
deri
ve
d
fr
om
palm
oi
l
,
sup
p
l
i
e
d by
L
i
on C
o
m
p
any
Sd
n. B
h
d. A
con
d
u
ct
i
v
e na
no
pa
rt
i
c
l
e
was used t
o
p
r
e
p
a
r
e t
h
e
PFAE
-
base
d n
a
no
fl
ui
ds. T
h
e
nan
o
p
art
i
c
l
e
was I
r
o
n
O
x
i
d
e (Fe
3
O
4
) wi
t
h
si
ze
range of 15
–
20 nm
.
Ol
ei
c
aci
d
was
use
d
as
t
h
e su
rfact
a
n
t
,
w
h
ere
b
y
0.
5 m
l
of
ol
ei
c aci
d
was m
i
xed
wi
t
h
t
h
e P
F
AE
oi
l
sam
p
l
e
s bef
o
re t
h
e
ad
d
ition
o
f
th
e Fe
3
O
4
n
a
nop
articles in
o
r
d
e
r to
m
o
d
i
fy the stab
ility o
f
t
h
e mix
t
u
r
es [1
7]. Th
e PFAE-b
ased
nan
o
f
l
u
i
d
s
we
re p
r
e
p
are
d
us
i
ng t
w
o
m
e
t
h
ods
i
.
e. M
e
t
h
od
I:
wei
g
ht
-
b
ased m
e
t
hod
[
19]
a
n
d
M
e
t
h
od
I
I
:
vol
um
e-fract
i
o
n m
e
t
hod
[
20]
.
In
gene
ral, for each m
e
thod,
there
are three
steps involve
d
in the pr
e
p
a
r
a
tion of na
nofluids, i.e. (1)
sam
p
l
e
wei
ghi
ng
,
(2
)
h
o
m
ogeni
zer t
r
eat
m
e
nt
a
n
d
(
3
)
vac
uum
pr
ocess
.
In
t
h
e
fi
rst
st
ep, t
h
e
na
no
pa
r
t
i
c
l
e
po
w
d
ers
were
wei
g
hed
usi
n
g
an anal
y
t
i
cal
bal
a
nce. F
o
r M
e
t
h
o
d
I, t
h
e
am
ount
o
f
Fe
3
O
4
n
a
nop
articles was
directly
m
easured using the analytical bala
nce [19], whereby the conce
n
tration of the
Fe
3
O
4
n
a
nop
articles
was ke
pt
fi
xed
at
0.
0
1
g/
l
.
On
t
h
e ot
her
han
d
,
f
o
r
M
e
t
h
od
I
I
,
t
h
e wei
g
ht
of
Fe
3
O
4
na
no
part
i
c
l
e
s
t
o
be use
d
was
det
e
rm
i
n
ed
o
n
t
h
e
basi
s of v
o
l
um
e
fract
i
on. The wei
g
ht
o
f
t
h
e
Fe
3
O
4
n
a
nop
articles is g
i
ven
b
y
th
e fo
llowing
fo
rm
ul
a [2
0]
,[
21]
:
100
nm
oil
nm
nm
PVF
V
W
(1
)
Whe
r
e
W
nm
is th
e weigh
t
of th
e Fe
3
O
4
n
a
nop
articles
(g
),
ρ
nm
is th
e d
e
n
s
ity o
f
Fe
3
O
4
na
no
pa
rt
i
c
l
e
s
(g/cm
3
),
V
oil
i
s
t
h
e
vol
um
e
of t
h
e P
F
AE
oi
l
sam
p
l
e
(
m
l) an
d
PV
F
nm
i
s
t
h
e
vol
um
e
fract
i
o
n
of t
h
e
Fe
3
O
4
nanoparticles (%). For exam
ple, the wei
ght
of t
h
e Fe
3
O
4
n
a
no
pa
rt
i
c
l
e
s requi
red
f
o
r
50
0
m
l
of i
n
s
u
l
a
t
i
on oi
l
i
s
0.
24
7
5
g
i
f
t
h
e
den
s
i
t
y
and
vol
um
e fract
i
on
o
f
t
h
e
nan
o
p
art
i
cl
es i
s
4.
95
g/
c
m
3 and 0
.
0
1
%,
respect
i
v
el
y
.
Tabl
e
1 s
h
o
w
s t
h
e c
once
n
t
r
at
i
o
n
o
f
t
h
e
Fe
3
O
4
nanoparticles and its
corresponding
weight for each 500 m
l
PFAE-
base
d
nan
o
fl
ui
ds
fr
om
M
e
t
hod
I a
n
d
M
e
t
h
o
d
II
. T
h
e
v
o
l
u
m
e
o
f
t
h
e
su
rfact
ant
i
s
al
so
s
h
o
w
n
i
n
Tabl
e
1.
The sec
o
n
d
st
ep i
n
vol
ves
ho
m
ogeni
zer t
r
ea
tm
ent
.
In t
h
i
s
st
ep, t
h
e sam
p
l
e
s were
hom
ogeni
ze
d i
n
or
der t
o
di
spe
r
se t
h
e F
e
3
O
4
n
a
nop
articles. It
is
worth
notin
g
h
e
re
th
at
th
is treatm
e
n
t
is ab
le to
break
up
n
a
nop
ar
ticle ag
gr
eg
ates and
ach
iev
e
w
e
ll-disp
er
sed
n
a
nopar
ticles [
1
9
]
,[22
],[2
3
]
.
The ho
m
o
g
e
n
i
zer was set at
50
%
f
o
r pul
se
-
m
ode
ope
rat
i
o
n (cy
c
l
e
)
a
n
d
5
0
% fo
r h
o
m
ogeni
zer p
o
we
r (
a
m
p
l
i
t
ude).
A
pr
o
b
e wi
t
h
a di
am
et
er
of
4
0
m
m
and
l
e
ngt
h
o
f
1
0
0
m
m
was used t
o
di
s
p
er
se t
h
e
Fe
3
O
4
n
a
no
p
a
rticles. Th
e th
ird
step
is t
h
e vacu
u
m
process
,
whe
r
e
b
y the
sam
p
les
we
re
placed i
n
a vac
u
um
ove
n to rem
ove
bubbles
, m
o
is
ture
and
gases t
h
at m
a
y
have
appea
r
ed during t
h
e
hom
ogenizer trea
tm
ent. The
sam
p
les were left in
th
e
v
a
cu
um
o
v
e
n
for at
least 72
hours at 70°C,
according to the proce
dure gi
ven i
n
[19]. The sa
m
p
les were
then stored in
s
ealed bea
k
ers
until
fu
rthe
r a
n
aly
s
is.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Co
mp
ara
tive
Stu
d
y
on
th
e AC
Brea
kdo
w
n Vo
lta
g
e
o
f
Pa
l
m
Fa
tty Acid Est
e
r .... (Moh
d Sa
fwan
Mo
hama
d
)
1
483
2.
2.
AC Breakdow
n Voltage Tes
t
It is kno
wn
t
h
at all electrical syste
m
s req
u
i
re testin
g
t
o
en
su
re th
at th
ey
will b
e
in
b
e
st
perfo
r
m
a
n
c
e
co
nd
itio
n
du
ri
n
g
o
p
eration
.
In
th
is case, it is i
m
p
o
r
tan
t
to
en
sure th
at in
su
lation
o
il is in
g
ood
cond
itio
n
du
ri
n
g
t
y
pi
cal
ope
rat
i
n
g co
n
d
i
t
i
ons o
f
t
h
e t
r
ansf
o
r
m
e
r.
The in
su
lation
o
i
l p
l
ays a v
ital ro
le to
i
n
su
late th
e
com
pone
nts in the trans
f
orm
e
r and it m
u
st be able to
withs
t
and the
r
m
a
l and electrical st
resses.
The
quality of
th
e in
su
latio
n
o
il is assessed
b
y
its resistiv
ity, p
e
rm
itt
iv
ity
an
d
AC
b
r
eakd
own
vo
ltag
e
.
Th
e
AC breakd
o
wn
v
o
ltag
e
test is on
e
o
f
th
e co
mm
o
n
test p
r
o
cedures
u
s
ed to
ev
alu
a
te th
e
q
u
ality o
f
t
h
e i
n
su
lation
o
il and
th
erefore, t
h
e
AC break
down vo
ltag
e
was ch
o
s
en
as th
e m
easu
r
e of th
e qu
ality o
f
th
e PFAE-b
ased
n
a
n
o
flu
i
d
s
in
th
is st
u
d
y
.
The AC
brea
k
d
o
w
n v
o
l
t
a
ge
of t
h
e P
F
A
E
-
b
ased na
n
o
fl
ui
d
s
was m
easured usi
ng M
e
gge
r OTS
6
0PB
po
rt
abl
e
oi
l
t
e
st
set
whi
c
h co
m
p
li
es wi
t
h
t
h
e ASTM
D
1
8
1
6
st
an
dar
d
t
e
st
m
e
t
hod. S
p
h
e
ri
cal
el
ect
rod
e
s wi
t
h
di
am
et
er of
36
m
m
were use
d
i
n
t
h
i
s
st
udy
and t
h
e ga
p d
i
st
ance bet
w
ee
n t
h
e el
ect
r
ode
s was 1
.
0 m
m
. Th
e
vol
um
e of eac
h P
F
AE
oi
l
sa
m
p
l
e
was kept
fi
xe
d at
50
0
m
l
. The AC
b
r
eakd
o
w
n
v
o
l
t
a
ge t
e
st
wa
s co
nd
uct
e
d
b
y
in
creasi
n
g
th
e vo
ltag
e
sup
p
l
y gradu
a
lly at a rate
o
f
0
.
5
kV/s un
til b
r
eak
down
occu
rs. A to
tal
o
f
40
b
r
eakd
own
vo
l
t
ag
es w
e
r
e
r
e
co
rd
ed
fo
r each PFAE
o
il sam
p
le an
d
th
e d
a
ta
we
re anal
ysed usi
ng sta
tistical
analysis.
Tabl
e 1. Pre
p
ar
at
i
on of
PF
AE
-
b
ase
d
nan
o
fl
ui
ds usi
n
g
m
e
t
h
o
d
I
a
n
d
m
e
t
hod
II
M
e
thod
Fe
3
O
4
nanopar
ticle
concentration
W
e
ight of Fe
3
O
4
nanopar
ticles per
500
ml
P
F
A
E
o
i
l
(g
)
Volu
m
e
of
surf
actant
Per
500 m
l
PFAE oil
(m
l)
M
e
thod I
0.
01 g/l
0.
0050
0.
25
M
e
thod I
I
0.
01%
0.
2457
0.
25
M
e
thod I
I
0.
02%
0.
4950
0.
25
M
e
thod I
I
0.
03%
0.
7425
0.
25
2.
3.
Weibull Sta
t
istica
l
Ana
l
y
s
is
Weibu
ll statisti
cal an
alysis was u
s
ed to
an
alyse th
e
AC
break
down vo
ltag
e
d
a
ta
for all sam
p
les. The
two-p
a
ram
e
ter
W
e
i
b
u
ll p
l
o
t
was u
s
ed
for
t
h
is
an
alysis.
Th
e b
r
eakd
own
p
r
ob
ab
ilities
were d
e
term
in
ed
u
s
i
ng
th
e fo
llowing
fo
rm
u
l
a [24
]
:
x
x
F
exp
1
(2
)
whe
r
e
F(x)
is th
e
p
r
ob
ab
ility o
f
th
e
AC
b
r
eak
down vo
ltage, x is th
e AC
b
r
eakd
own
v
o
l
tag
e
,
β
is th
e
sh
ap
e
param
e
t
e
r and
η
i
s
t
h
e scal
e
param
e
t
e
r. Th
e W
e
i
bul
l
pl
ot
was pl
ot
t
e
d b
a
sed o
n
t
h
e est
i
m
a
t
i
on of t
h
e
shape
param
e
t
e
r, scal
e param
e
t
e
r and c
o
r
r
el
at
i
on
coefficient. T
h
e correlation coefficient,
ρ
is a
m
easure of how
well th
e lin
ear
m
o
d
e
l fits th
e
d
a
ta b
e
tween
t
h
e m
e
d
i
an
ranks. T
h
e correla
tion coe
fficient
has a value
be
tween
–
1
t
o
+1
. In
g
e
n
e
ral, a correlatio
n
co
efficient clo
s
e to
+1
in
d
i
cates th
at t
h
ere is a st
rong
po
sitiv
e co
rrelatio
n
b
e
tween
two
variab
les,
wh
ich is ev
i
d
en
t
from
th
e up
ward
t
r
end
i
n
th
e gr
ap
h.
On
th
e o
t
her
hand
, a co
rrelatio
n
coefficient close to
–1 i
ndica
tes that
there i
s
a strong
nega
tiv
e co
rrelation
between
two v
a
riab
les,
wh
i
c
h
is
also
ev
id
en
t fro
m
th
e d
o
wn
ward trend
i
n
t
h
e graph
.
Plo
ttin
g
th
e
Weib
u
ll p
r
ob
ab
il
ity p
l
o
t
is
easy
fo
r tho
s
e who are familiar
with
lin
ear p
l
ots an
d
b
a
sic
al
geb
r
a.
The
e
quat
i
o
n
use
d
t
o
pr
o
duce
a l
i
n
ear
pl
ot
i
s
gi
ven
by
Eq
uat
i
o
n
(
3
)
,
whe
r
e
y
i
s
t
h
e
d
e
pe
nde
n
t
vari
a
b
l
e
,
x
i
s
t
h
e i
nde
pe
nde
nt
vari
a
b
l
e
,
m
i
s
t
h
e sl
ope
an
d
b
is th
e in
tercep
t.
b
mx
y
(3
)
Eq
uat
i
o
n
(
2
)
w
a
s m
a
ni
pul
at
ed
i
n
or
der
t
o
gi
v
e
t
h
e
fol
l
o
wi
n
g
eq
uat
i
on:
ln
x
y
(4
)
eq
u
a
tion
(4) is a lin
ear equatio
n
,
wh
ereby
β
i
s
t
h
e sl
ope an
d
β
ln
(
η
)
is th
e in
tercep
t.
The sl
ope can
be
d
e
term
in
ed
from
th
e lin
ear
p
l
o
t
,
wh
ich
in turn
,
g
i
v
e
s th
e v
a
l
u
e
o
f
t
h
e sh
ap
e p
a
ram
e
ter,
β
. It sh
all b
e
no
ted th
at
th
e x
-
ax
is rep
r
esen
ts th
e AC b
r
eakd
own
vo
ltag
e
, wh
ich
is
arra
nge
d f
r
om
t
h
e l
o
west
t
o
t
h
e hi
g
h
est
val
u
e. Th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 4
,
Au
gu
st 2
016
:
14
81
–
1
488
1
484
y
-
axi
s
re
prese
n
t
s
t
h
e
m
e
di
an r
a
nk
. The m
e
di
an ra
nk i
s
det
e
rm
i
n
ed for eac
h AC
brea
kd
o
w
n
v
o
l
t
a
ge val
u
e a
n
d
i
s
gi
ve
n
by
t
h
e
fol
l
o
wi
n
g
e
q
ua
t
i
on
[2
4]
:
4
.
0
3
.
0
N
j
MR
(5
)
whe
r
e
j
is th
e
sequ
en
ce
o
f
d
a
ta fro
m
1
to
N
and
N
is the nu
m
b
er o
f
d
a
ta. For ex
am
ple, if the num
b
er
of
AC
brea
k
d
o
w
n
vol
t
a
ges re
co
rde
d
pe
r e
xpe
ri
m
e
nt
i
s
4
0
(w
hi
ch
i
s
t
h
e case
i
n
t
h
i
s
st
u
d
y
)
, t
h
e
n
N
= 40
. Thus, th
e
med
i
an
rank
need
s t
o
b
e
d
e
term
in
ed
40
tim
e
s
, i.e.
one
m
e
dian rank for
eac
h AC
b
r
eak
dow
n
vo
ltag
e
wher
e
j
=
1
,
…,40
. Th
e
med
i
an
rank
is in
th
e form
o
f
a
ratio
, w
ith
a value
greater than 0 a
n
d less tha
n
1.
T
h
e scale
param
e
ter,
η
, i
ndicates the
spread
of t
h
e
distribution a
n
d it can
be
use
d
to
stretch
or s
queeze the graph [24].
In th
is case, the prob
ab
ility o
f
th
e AC
breakd
o
wn
v
o
ltag
e
wh
en
x=
η
=
β
i
s
show
n in
Equatio
n
(
6
)
.
%
2
.
63
632
.
0
1
exp
1
exp
1
x
x
F
(6
)
In
ge
neral
,
t
h
e
sha
p
e
param
e
ter ca
n
be
det
e
r
m
i
n
ed o
n
ce t
h
e
co
rrel
a
t
i
o
n c
o
effi
ci
ent
a
n
d m
e
di
an
ra
nk
are
k
nown. The scale p
a
ram
e
ter can b
e
b
e
st
esti
m
a
ted
after p
l
o
tting
t
h
e
g
r
ap
h.
3.
R
E
SU
LTS AN
D ANA
LY
SIS
3.
1.
AC
B
reak
d
ow
n V
o
l
t
ag
e
The
di
st
ri
b
u
t
i
o
n o
f
t
h
e
AC
brea
k
d
o
w
n v
o
l
t
a
ge fo
r t
h
e
vi
r
g
i
n
P
F
AE
oi
l
,
PF
AE-
b
as
ed na
n
o
fl
ui
d
pre
p
are
d
u
s
i
n
g
M
e
t
hod I a
nd
PFAE
-
base
d n
a
no
fl
ui
d pre
p
a
r
ed
usi
n
g M
e
t
h
od I
I
are s
h
o
w
n i
n
Fi
g
u
res
1 (a),
(b
)
and
(c)
,
res
p
ec
t
i
v
el
y
.
It
can b
e
seen f
r
om
Figu
re 1
(a) t
h
at
t
h
e l
o
west
A
C
brea
kd
o
w
n
vol
t
a
ge
fo
r t
h
e
vi
rg
i
n
PFAE
oil is 9
kV at instance
36
whereas t
h
e
highest AC
br
eakd
o
w
n
v
o
l
t
a
ge i
s
3
0
kV
at instance 5.
Si
milarly,
it can be seen from
Figure
1 (b) that the
lowest AC
br
eakdo
wn
vo
ltag
e
fo
r
th
e PFA
E
-
b
ased
n
a
n
o
f
l
ui
d
pre
p
are
d
usi
n
g
M
e
t
h
o
d
I i
s
1
1
k
V
at
i
n
st
a
n
c
e
34
an
d
37
, w
h
i
l
e
t
h
e hi
ghe
st
AC
b
r
ea
kd
o
w
n v
o
l
t
a
ge i
s
3
3
kV
at
i
n
st
ance
4. It
can be see
n
f
r
om
Fi
gure
1
(c) t
h
at
t
h
e
hi
ghe
st
AC
b
r
ea
kd
o
w
n
v
o
l
t
a
ge
of t
h
e
PF
AE-
b
ase
d
nan
o
f
l
u
i
d
co
nt
ai
ni
ng
0.
01
,
0.
02
a
n
d
0
.
0
3
%
Fe
3O
4
na
no
p
a
rt
i
c
l
e
s pr
epa
r
ed
usi
n
g
M
e
t
h
od
I
I
i
s
3
2
,
3
8
and
35
kV, respectivel
y, whereas
the
lowe
st A
C
b
r
eak
dow
n
vo
ltage
is
8, 1
7
and
15
k
V
,
resp
ectively.
3.
2.
Da
ta
A
n
al
ysi
s
Fi
gu
re 2 (a
) sho
w
s t
h
e t
w
o-
param
e
t
e
r Wei
b
ul
l
pl
ot
of t
h
e vi
rgi
n
PFAE
oi
l
and
PFAE
-
base
d
n
a
nof
lu
id pr
epar
ed u
s
i
n
g Meth
od
I and
Figu
r
e
2
(b
)
show
s th
e two
-
p
a
r
a
m
e
ter
W
e
i
b
ull p
l
o
t
of
t
h
e
PFA
E-
base
d na
n
o
fl
ui
ds p
r
e
p
are
d
usi
ng M
e
t
h
o
d
I
I
.
The
AC
b
r
eak
do
w
n
v
o
l
t
a
ge
dat
a
f
o
r t
h
e
vi
r
g
i
n
P
F
AE
oi
l
a
r
e al
s
o
sup
e
r
pos
ed i
n
Figu
re
2 (a
) a
nd
(
b
) as
refe
r
e
nce. Ta
ble
2 shows t
h
e
values
of t
h
e sha
p
e pa
ram
e
ter, scale
param
e
t
e
r and cor
r
el
at
i
on coe
ffi
ci
ent
f
o
r t
h
e vi
r
g
i
n
PF
AE oi
l
,
PFAE
-
base
d
nan
o
f
l
u
i
d
s
pre
p
are
d
usi
ng M
e
t
h
o
d
I an
d Meth
od
II
b
a
sed
on
th
e two-p
a
ram
e
te
r
Weibu
ll p
l
o
t
s
in Fig
u
res
2
(a)
a
n
d (b
). It can be
see
n
fr
om
the
Table 2 t
h
at the correlation
coefficient for each sam
p
le
i
s
close to
1 sugge
sting t
h
ere
is a strong,
positive
co
rrelatio
n
b
e
t
w
een
th
e W
e
i
b
u
ll
prob
ab
ility
an
d AC b
r
eak
d
o
w
n
vo
ltag
e
.
Based
on
th
e
lin
ear
lin
es
i
n
Fig
u
res
7 (a) a
nd
(b
),
the shape
par
a
m
e
ter,
β
for
t
h
e vi
r
g
i
n
PF
AE oi
l
an
d P
F
AE
-bas
ed
na
no
fl
ui
d wi
t
h
a
Fe
3
O
4
nan
o
p
art
i
c
l
e
conce
n
t
r
at
i
o
n o
f
0.0
1
g/
l
,
0.
01
%, 0.
0
2
% an
d
0.0
3
% i
s
4.
7
8
, 4
.
4
7
,
6.
5
3
and
7.
03
res
p
ec
t
i
v
el
y
.
The sha
p
e pa
ra
m
e
ter indeed affects the shap
e
of t
h
e
Wei
b
ul
l
pl
ot
s.
The po
p
u
l
a
t
i
on
of
β
< 1 exhi
bi
t
a
p
r
ob
ab
ility th
at d
ecrease with th
e AC b
r
eakd
o
wn
vo
ltag
e
, th
e p
opu
latio
n
o
f
β
= 1
h
a
v
e
a co
n
s
tan
t
probab
ility
and t
h
e
po
p
u
l
a
t
i
on o
f
β
> 1
h
a
v
e
a p
r
ob
ab
ility th
at
in
crease with
th
e AC breakdo
wn
vo
ltag
e
. Tab
l
e 3
sh
ows
t
h
e
AC
b
r
eak
d
o
w
n
v
o
l
t
a
ge
f
o
r
t
h
e vi
r
g
i
n
PF
AE oi
l
an
d
P
F
A
E-
base
d nan
o
fl
ui
ds pre
p
a
r
e
d
usi
n
g
M
e
t
h
o
d
I
an
d
Meth
od
II based
o
n
t
h
e
Weib
u
ll
p
r
o
b
a
b
ilities o
f
Figu
re
s 7
(a)
an
d
(b
). Based
o
n
t
h
e Weibu
ll
p
r
ob
ab
ility
o
f
63
.2
%, t
h
e A
C
break
d
o
w
n
vol
t
a
ge i
s
t
h
e
l
o
west
fo
r t
h
e PFAE
-
ba
sed
nan
o
fl
ui
d co
nt
ai
ni
n
g
0.
0
1
%
Fe
3
O
4
n
a
nop
articles co
m
p
ared
to
o
t
her sam
p
les, with
a v
a
lu
e
o
f
17.45
kV
. In
co
n
t
r
a
st, th
e A
C
b
r
eak
dow
n
vo
ltag
e
is
t
h
e hi
ghest
fo
r
t
h
e PF
AE
-ba
s
ed
nan
o
fl
ui
d
p
r
epa
r
ed
usi
n
g
M
e
t
h
o
d
I
I
c
o
n
t
ai
ni
ng
0.
0
2
%
Fe
3
O
4
nanop
articles,
wi
t
h
a val
u
e
o
f
3
0
.
35
kV
. I
n
t
e
rest
i
ngl
y, ev
en
a slig
h
t
in
crease in
th
e Fe
3
O
4
na
no
part
i
c
l
e
conce
n
t
r
at
i
o
n has a
d
r
am
atic ef
f
ect on
th
e A
C
br
eak
dow
n vo
ltage of
th
e PFA
E
-b
ased
n
a
no
f
l
u
i
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Co
mp
ara
tive
Stu
d
y
on
th
e AC
Brea
kdo
w
n Vo
lta
g
e
o
f
Pa
l
m
Fa
tty Acid Est
e
r .... (Moh
d Sa
fwan
Mo
hama
d
)
1
485
(a)
(b
)
(c)
Figu
re 1.
Distri
butio
n o
f
the
A
C
brea
k
d
o
w
n v
o
ltage fo
r (a)
t
h
e vir
g
in PF
A
E
oil, (b
)
t
h
e
P
F
AE
-bas
ed
n
a
nof
lu
id pr
epar
ed u
s
i
n
g Meth
od
I
.
, an
d (
c
)
th
e PFA
E
-
b
ased
n
a
nof
lu
id
s prep
ar
ed
u
s
ing
Meth
od
II
Table
2.
Sha
p
e
pa
ram
e
ter, scale param
e
ter and corre
l
a
t
i
o
n coef
fi
ci
ent
obt
ai
ned fr
om
t
h
e t
w
o
-
pa
ram
e
t
e
r
wei
b
ul
l
pl
ot
s
f
o
r
vi
r
g
i
n
oi
l
a
n
d
pfae
-ba
s
ed
n
a
no
fl
ui
ds
pre
p
ared
usi
n
g
m
e
tho
d
I a
n
d m
e
t
hod
I
I
Para
m
e
ter
Virgin
PFAE
oil
PFAE
-
based nanofluid
(
M
ethod I
)
PFAE
-
based nanofluids
(
M
ethod I
I
)
0.
01 g/l Fe
3
O
4
0.
01%
Fe
3
O
4
0.
02%
Fe
3
O
4
0.
03%
Fe
3
O
4
β
4.
78
4.
47
3.
38
6.
53
7.
02
η
19.
26
23.
84
17.
45
30.
35
27.
14
ρ
0.
96
0.
99
0.
93
0.
97
0.
97
Tab
l
e
3
.
Weibu
ll prob
ab
ility
for
AC break
d
o
w
n
vo
ltag
e
o
f
all sam
p
les b
a
sed
on
th
e two
-
p
a
ram
e
ter wei
b
u
ll
W
e
ibull
probability
(%)
AC br
eakdown voltage (
k
V)
Virgin PFAE
oil
PFAE
-
based nanofluid
(
M
ethod I
)
PFAE
-
based nanofluids
(
M
ethod I
I)
0.
01 g/l Fe
3
O
4
0.
01%
Fe
3
O
4
0.
02%
Fe
3
O
4
0.
03%
Fe
3
O
4
1
7.
36
8.
52
4.
48
15.
00
14.
10
5
10.
34
12.
26
7.
24
19.
26
17.
78
10
12.
03
14.
41
8.
97
21.
50
19.
70
50
17.
84
21.
97
15.
66
28.
70
25.
76
63.
2
19.
26
23.
84
17.
45
30.
35
27.
14
It appea
r
s that
a proper
perce
n
tage
of Fe
3
O
4
is im
p
o
r
tan
t
t
o
ach
i
ev
e th
e
h
i
gh
est im
p
r
ove
m
e
n
t
in
th
e
AC break
down vo
ltag
e
lev
e
l.
Th
e ad
d
ition
of 0.01
g
/l of Fe
3
O
4
n
a
no
p
a
rticles in
to
th
e PFAE
o
il in
creases its
AC b
r
eak
down
vo
ltag
e
b
y
2
3
.8
%.
Th
is
is ob
v
i
o
u
s wh
en
t
h
e
ad
d
ition
of 0
.
0
2
%
Fe
3
O
4
nan
o
p
a
rticles into
th
e
PFAE
o
il
i
n
crease
th
e AC breakdo
wn
vo
ltag
e
b
y
57
.6
%
at
a Weibu
ll
p
r
ob
ab
ility
o
f
6
3
.2%. Ho
wev
e
r,
a
furthe
r increas
e in the a
m
ount of Fe
3
O
4
(
0
.
0
3%)
wo
ul
d see
a bi
t
reduct
i
o
n
i
n
t
h
e im
prov
em
ent
of brea
k
d
o
w
n
v
o
ltag
e
lev
e
l.
The c
o
nductive na
nopa
rticles act as elect
ron sca
v
e
n
gers in t
h
e i
n
sul
a
tion
oil unde
r
electrical
stresses,
which is conve
r
t the
fast electrons i
n
to sl
ow
ne
gat
i
v
el
y
char
ge
d
p
a
rt
i
c
l
e
s [1
4]
,[
2
5
]
.
I
n
ot
her
w
o
rds
,
i
t
takes a long tim
e for the stream
er proces
s to brea
kdown
unde
r electrical
stress. Figure
3 shows an illustration
o
f
t
h
e stream
er pro
cess
for t
h
e in
su
lation
oil with
and
with
ou
t co
ndu
ctive n
a
n
o
p
a
rticles. It can
b
e
seen
th
at
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 4
,
Au
gu
st 2
016
:
14
81
–
1
488
1
486
the stream
er is
in direct conta
c
t be
tween
t
h
e h
i
gh
vo
ltag
e
electro
d
e
and
gro
und
fo
r t
h
e insu
latio
n
o
il with
ou
t
co
ndu
ctiv
e
n
a
n
o
p
a
rticles and
th
erefore,
t
h
e stream
er takes a s
h
orter time
to
br
eak
dow
n.
In
co
n
t
r
a
st, th
e
str
eam
er
tak
e
s a lo
ng
er
time to
br
eakdow
n
f
o
r
th
e i
n
su
latio
n
o
il with
co
ndu
ctiv
e n
a
nop
ar
ticles u
pon
application of high voltage because
of
electron
sca
v
e
nge
r action.
Figure
4 shows the effect of
Fe
3
O
4
na
n
opa
r
t
i
c
l
e
concent
r
a
t
i
on o
n
t
h
e
AC
brea
k
d
o
w
n v
o
l
t
a
ge of t
h
e
PFAE-b
a
sed
nan
o
flu
i
d
at a
Weibu
ll p
r
ob
ab
ility o
f
63
.2
%. It can
b
e
seen
th
at the add
itio
n
o
f
0.01% Fe
3
O
4
n
a
nop
articles in
to
t
h
e PFAE o
il d
e
creases
th
e AC
b
r
eakd
own
vo
ltag
e
, wh
ich
co
n
t
rad
i
cts wh
at is in
itially
expect
e
d
i
n
t
h
i
s
st
u
d
y
.
Di
ffe
rent
c
onc
ent
r
at
i
ons
o
f
na
n
opa
rt
i
c
l
e
s have
be
en
used
i
n
pre
v
i
o
us st
udi
es
[
13]
,
[2
6]
,[
2
7
]
t
h
e resul
t
s
sh
ow t
h
at
som
e
con
cent
r
at
i
o
ns d
o
not
i
m
prove
t
h
e AC
b
r
eak
do
w
n
v
o
l
t
a
ge
of t
h
e
d
i
electric liq
u
i
d
.
Ind
e
ed
, th
e resu
lts
o
f
t
h
is stu
d
y
ind
i
cat
e th
at 0
.
01
% is no
t th
e id
eal
Fe
3
O
4
na
n
opa
rt
i
c
l
e
conce
n
t
r
at
i
o
n
f
o
r
PF
AE-
b
ase
d
na
no
fl
ui
ds.
(a)
(b
)
Fi
gu
re
2.
Tw
o
-
param
e
t
e
r
W
e
i
bul
l
pl
ot
of
t
h
e
AC
brea
k
d
o
w
n
vol
t
a
ge
f
o
r
(a
) t
h
e
vi
r
g
i
n
PF
AE
oi
l
an
d
PF
AE-
base
d
nan
o
fl
ui
d
pre
p
are
d
usi
n
g
M
e
t
h
od
I,
a
n
d
(
b
)
t
h
e
vi
r
g
i
n
P
F
AE
oi
l
a
n
d P
F
AE
-
b
ase
d
nan
o
f
l
u
i
d
p
r
ep
ared
usi
n
g M
e
t
h
o
d
I
I
Fig
u
re
3
.
Illu
st
ratio
n of streamer p
r
o
cess in
in
su
lation
o
il with
an
d withou
t cond
u
c
ti
v
e
n
a
nop
articles
4.
CO
NCL
USI
O
N
The AC
b
r
eak
do
w
n
v
o
l
t
a
ge of PF
AE oi
l
s
m
i
xed wi
t
h
Fe
3
O
4
nan
o
p
art
i
c
l
e
s at
di
ffere
nt
conce
n
t
r
at
i
o
n
s
has bee
n
i
n
ves
t
i
g
at
ed i
n
t
h
i
s
st
udy
. T
w
o m
e
t
h
o
d
s are
use
d
t
o
p
r
e
p
are t
h
e PF
AE-
b
ase
d
na
n
o
fl
ui
ds,
nam
e
ly
Meth
od
I an
d
Meth
od
II. Meth
od
I is
b
a
sed
o
n
th
e
wei
g
ht
o
f
t
h
e
na
no
part
i
c
l
e
s
whe
r
eby
t
h
e
PF
AE
-base
d
n
a
noflu
id is
p
r
ep
ared
b
y
m
i
x
i
n
g
th
e PFAE oil with
0
.
0
1
g
/l of Fe
3
O
4
na
no
part
i
c
l
e
s. M
e
t
h
od
I
I
i
s
bas
e
d
on
t
h
e
v
o
l
u
m
e f
r
actio
n
of
th
e
n
a
nopar
ticles, w
h
er
eb
y th
e PFA
E
-
b
ased
n
a
no
f
l
u
i
ds ar
e pr
epar
ed
b
y
mix
i
n
g
th
e
PFAE
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Co
mp
ara
tive
Stu
d
y
on
th
e AC
Brea
kdo
w
n Vo
lta
g
e
o
f
Pa
l
m
Fa
tty Acid Est
e
r .... (Moh
d Sa
fwan
Mo
hama
d
)
1
487
o
il at th
ree d
i
fferen
t
con
c
en
tration
s
of Fe
3
O
4
nan
o
p
art
i
cl
es (0.
0
1
,
0.
02 a
nd
0.
03%
). The P
F
AE
-
b
ase
d
nan
o
f
l
u
i
d
s a
r
e
m
i
xed usi
ng a
l
a
bo
rat
o
ry
ho
m
ogeni
zer t
o
e
n
su
re t
h
at
t
h
e
nan
o
p
art
i
c
l
e
s a
r
e wel
l
-
di
spe
r
s
e
d i
n
th
e PFAE
o
ils. Th
e resu
lts
are an
alysed
u
s
ing
Weibu
ll
statistical an
alysis an
d
it is foun
d th
at there is
en
h
a
n
cem
en
t in
th
e AC
break
down vo
ltag
e
o
f
th
e
PFAE
-b
ased
n
a
no
fl
u
i
d
s
with
the
excep
tio
n
of on
e sam
p
le
pre
p
are
d
usi
n
g
M
e
t
hod
II (c
o
n
t
a
i
n
i
n
g 0
.
0
1
%
Fe
3
O
4
na
noparticles). The pre
s
ence of Fe
3
O
4
n
a
nop
articles slo
w
s
down the
strea
m
er process
in
the PFAE
oil
due to the electron sca
v
enger a
c
tion.
Figure
4
.
E
ffec
t
of
Fe
3
O
4
na
n
opa
rt
i
c
l
e
co
nce
n
t
r
at
i
o
n
on
t
h
e
AC
b
r
ea
kd
o
w
n
v
o
l
t
a
ge
of
PF
AE-
b
ase
d
na
no
fl
ui
ds
at
a W
e
i
b
u
ll p
r
o
b
a
b
ility
o
f
63
.2
%
ACKNOWLE
DGE
M
ENTS
The a
u
thors
gratefully ac
knowle
dge
the
technical
as
sistance
provide
d
by th
e
staff
at Faculty of
Electrical En
gin
eering
,
Un
iversiti Tekn
ik
al
Malaysia
Me
lak
a
. Th
is
wo
rk
was sup
p
o
r
ted
in
p
a
rt b
y
the
M
i
ni
st
ry
of
Ed
ucat
i
on M
a
l
a
y
s
i
a
unde
r
t
h
e R
e
sear
ch Acc
u
l
t
u
rat
i
on
Gra
n
t
S
c
hem
e
(Grant
no
.:
R
AGS/
1
/
2
0
14/
TK
03/
F
K
E/
B
0
00
5
5
) a
n
d
Fun
d
am
ent
a
l
R
e
search
Gra
n
t
Sc
hem
e
(Gran
t
no.:
FR
GS/
1
/
2
01
5/
TK
04/
F
K
E/
0
3
/
F
0
0
2
6
2
).
REFERE
NC
ES
[1]
W.
Z
i
ome
k
,
et
a
l
.
,
“
H
igh Voltag
e
Power Tr
ansform
e
r Insulation
Design,”
El
ectr
i
cal Ins
u
la
tion C
onfer
enc
e
(
E
IC)
,
pp. 211–215
, 20
11.
[2]
P.
K.
Pa
ndey
,
et al
, “
ANF
IS
Based Approach
to
Es
tim
ate R
e
m
n
a
n
t Li
f
e
of Power
transform
e
r b
y
Predicting
Furan
Contents,”
Int
.
J.
El
ectr
i
cal
.
Com
puter
Eng
i
neer
in
g
, vol/issue: 4(4)
, pp
. 463-470
, 2
014.
[3]
P. Soundiraraju,
et
al
, “Wavelet Transforms based Po
wer tr
ansf
ormer Protectio
n from Magnetic Inrush Curr
ent,”
Telkomnika
Indo
nesia Journal
of
Ele
c
trica
l
Eng
i
n
eering
, vol/issue: 14(3), pp. 381-
387, 2015
.
[4]
W
.
Ziom
ek, “
T
rans
form
er elec
t
r
ica
l
ins
u
lat
i
on,
”
IEEE Transactions on Diel
e
c
tric and E
l
ec
t
r
ical Insulation
,
vol/issue: 19(6), pp.
1841–1842
, 2012.
[5]
M. S. Naidu,
et al.
, “
High-Volta
ge Eng
i
neering
5e
,” 5e ed
, Ind
i
a, McGraw Hi
ll Education, pp. 1–
468, 2013
.
[6]
M.
Firda
u
s bin A.
Ka
ma
l,
“T
he
Effect of
Ele
c
t
r
ica
l
Ageing on
Elec
tri
cal P
r
op
erti
es
of P
a
lm
F
a
tt
y
A
c
id Es
te
r
(PFAE) and FR3 as High
Voltag
e
Insulation,”
Uni
versiti
Teknolog
i
Malaysia
, pp
. 1
–54, 2013
.
[7]
N. S. Murad,
et al.
, “A Stud
y
o
n
Palm Oil Moisture Ab
sorption Level
and Voltage Breakdown,”
Annual Report
Conference on
Electrical Insu
lation and Dielectric Ph
enomena,
p
p
. 925–928
, 201
3.
[8]
N.
Azis,
et al.
,
“
S
uitabi
lit
y
o
f
palm
bas
e
d oi
l as d
i
el
ec
tric
i
n
sulating
fluid
i
n
transform
e
rs,”
J.
E
l
ec
tr
.
Eng
.
Technol.
, vo
l. 9,
pp. 662–669
, 20
13.
[9]
H.
Kojima,
et al.
, “
C
harge behavi
or in Palm
Fatt
y Acid Ester Oil
(PFAE) / pressboard com
posite in
sulation s
y
st
em
under voltage ap
plication
,
”
Conf. Rec. IEEE In
t. S
y
mp. Electr. Insul.
, pp. 419–423, 2014.
[10]
M. Corporation,
“Features of
Eco
-
Fr
iendly
Transf
ormers Using Palm Fatty
Ac
id E
s
ter ( PFAE )
,
a
New Vegetable-
Based Insula
tion
Oil,
”
Mid
e
n R
e
view
,
no
. 163
, p
p
. 39–45
, 2015
.
[11]
M.
T.
Che
n
,
et al.
, “Effect of n
a
noparticles on
the dielectric strength of
ag
ed tr
ansformer oil,”
Annu. Rep. -
Conf.
Ele
c
tr.
Insul.
Di
ele
c
tr. Phenomena, CEIDP
,
pp
. 664–667,
2011
.
[12]
H. Jin,
et a
l
.
, “AC breakdown voltag
e
and viscosity
of
mineral oil based SiO2
nanofluids,”
An
nu. Rep
.
- Con
f
.
Ele
c
tr.
Insul.
Di
ele
c
tr. Phenomena, CEIDP
, pp
.
902–905, 2012
.
[13]
Y. Z.
Lv
,
et al.
,
“
R
ecent
P
r
ogre
s
s
in Nanofluids
Bas
e
d on
Trans
f
orm
e
r Oil
: P
r
e
p
arat
ion and
El
e
c
tri
cal
Ins
u
lat
i
o
n
Properties,”
Electrical Insulation
Magazine, I
E
EE,
vol/issue: 30(5)
, pp
. 23–32
, 201
4.
[14]
L. Yuzh
en,
et al.
, “Nanopar
ticle
Effect on Diel
ectric Br
eakdown
Strength of
Tran
sformer Oil-Based Nanoflu
ids,”
Annu.
Rep
.
-
Co
nf.
Electr. Insu
l.
Dielectr. Pheno
m
ena, CEID
P
, p
p
. 680–682
, 201
3.
[15]
H.
U.
Zhi-feng,
et al.
, “Th
e
rmal
Aging Properties of Tran
sformer Oil-Based
TiO
2 Nanofluids,”
Proc. - I
E
EE Int.
Conf. Dielectr. Liq.
, pp. 1–4, 201
4.
[16]
Y. Zhou,
et a
l
.
, “Eff
ect of
n
a
noparticles on
electrical ch
ar
act
eristics of tr
ansformer oil-b
a
sed nanoflu
ids
impregnated
pressboard,”
Conf
.
Rec
.
I
EEE
Int
.
S
y
m
p
. E
l
e
c
tr
.
Ins
u
l.
, pp. 650–653, 2012.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 4
,
Au
gu
st 2
016
:
14
81
–
1
488
1
488
[17]
D. E. a Mansour,
et al.
, “Effect of titan
i
a nanoparti
c
les on the diel
ectr
i
c pr
operti
e
s of transform
e
r oil-based
nanofluids,”
Ann
u
. R
e
p.
- Con
f
.
E
l
ec
tr
. Ins
u
l
.
Di
el
ectr
.
Ph
enomena
,
CEIDP
, pp
. 29
5–298, 2012
.
[18]
D. Martin
et al.
, “Statistical
analy
s
is of
the AC
breakdown
voltages of ester b
a
sed transformer oils,”
IE
EE T
r
ans.
Diele
c
tr
.
El
ectr
.
Ins
u
l.
, vol. 15, p
p
. 1044–1050
, 2
008.
[19]
W.
X.
Si
ma
,
et al.
, “Prepar
a
tio
n of Thr
e
e Tran
sformer Oil-Base
d Nanoflu
ids and Comparison
of Their Impulse
Breakdown Char
act
eristics,”
Nan
o
sci. Nano
te
chn
o
l. Le
tt.
, v
o
l. 6,
pp. 250–256
, 20
14.
[20]
P.
Krishna Kumar,
et a
l
.
, “Investigation
on mixed insulating f
l
uids
with nano
flu
i
ds and
antiox
i
d
a
nts,”
Adv
.
El
ect
r.
Eng. (
I
CAEE)
,
2
014 Int. Conf.
, p
p
. 1–4
, 2014
.
[21]
M.
H.
Cha
n
g,
et al.
, “Prepar
a
tion
of copper ox
ide nanopartic
les an
d its application
in nanoflu
id,”
Powder Technol.
,
vol. 207
, pp
. 378
–386, 2011
.
[22]
H. J
i
n, “
D
iel
ect
ric S
t
reng
th an
d Therm
a
l Co
n
ductivity
of Mineral Oil bas
e
d
Nanofluids,”
D
e
lft
Univ
ersity
of
Technology
, pp
.
1–170, 2015
.
[23]
X.
Q.
Wang,
et al.
, “A review o
n
nanofluids - Part
II: Exp
e
riments and applications,”
Brazilian
J. Chem. Eng
.
,
vol/issue:
25(04)
, pp
. 631–648
, 2
008.
[24]
R. Corpora
tion
,
“
L
ife Da
ta An
al
ysis Re
feren
c
e
,
”
ReliaSo
ft Corporation,
pp
. 1–43
8, 2015
.
[25]
J.
G.
Hwang,
et al.
, “Electron
Scavenging b
y
Conductive Nan
opartic
les in Oil Insulated Power Transformers,”
Ele
c
trostati
cs Jo
int Conf
erenc
e
,
pp.
1–12
,
2009
.
[26]
S. Grzy
bowski,
et al.
, “Preparation of
a Veg
e
tab
l
e Oil-B
a
se
d Nanofluid
an
d Investigation
of Its,”
Electrical
Insulation Maga
zine, I
E
EE
, vol/issue: 28(5), pp.
43–50, 2012
.
[27]
R.
K.
A.
Cavallini,
et al.
, “Investigations on the effect of
na
nopa
rt
ic
l
e
s i
n
mine
r
a
l
oi
l
,
”
IEEE, Annual
Report
Conference on
Electrical Insu
lation and Dielectric Ph
enomena,
p
p
. 695–698
, 201
4.
BIOGRAP
HI
ES
OF AUTH
ORS
M
o
hd Safw
an
M
o
hamad
rece
i
v
ed th
e Diplom
a in
El
ec
tric
al
E
ngineer
ing and
Bache
l
or d
e
gree
in El
ectrical
Eng
i
nering
(Industri
a
l power)
from
Un
iversiti T
e
kn
i
k
al
Malay
s
ia Melaka (UT
e
M) in
201
1 and
2014, respectiv
ely
.
He is
curr
ently
pur
suing his
MSc in
Electrical Eng
i
neer
ing
at
UT
eM
.
His
r
e
s
e
arch fo
cus
e
s
on
nano-fluid
insulation for
high
vol
tage application.
Hiday
at
Z
a
inuddin
receiv
e
d
his
Bach
elor
of
Engin
eering
in
El
ectr
i
c
a
l
from
the
Univers
i
ti
T
e
kno
logi Malay
s
ia (UTM)
in
2
003.
He obtained his MSc in
Elec
trical Power
En
gineer
ing with
Business from the University
o
f
S
t
rathcly
d
e, Glasgow in 2005
.
He r
e
ceived
his
PhD degree
at
the Univ
ersity
o
f
Southampton,
United Kindom
in
2013.
He has been
as
academic staf
f
of
the
Universiti
T
e
knikal Malay
s
ia Mela
k
a
(U
T
e
M)
since 2003
and
at
present is
the senior
lecturer
in
F
acult
y of
El
ec
t
r
ica
l
Eng
i
neer
in
g at
the un
ivers
i
t
y
. H
e
is
th
e
Lab He
ad for
th
e High
V
o
lt
age
Engineering Res
earch
Labor
ator
y
,
Univ
ersiti
T
e
knikal Ma
la
ysia
Melaka
(UT
e
M
)
. His resea
r
ch
inter
e
sts inc
l
ude
HV
equipm
ent
and insul
a
tion
c
ondition
m
onitor
i
ng, f
a
i
l
ure
ana
l
ysis,
and
power
s
y
stem protectio
n coordin
a
tion
.
Shar
i
n Ab
Ghani
rec
e
iv
ed
the
BEng.
(Hons) d
e
gree
in
E
l
ec
tri
c
al
Engin
eering
f
r
om
Universiti
T
e
kn
ikal Malaysia Melaka (UT
e
M) in 2008
an
d
the MEng. Degree in
El
ectrical Engin
eer
ing
from Universiti
T
e
naga Nasion
al, Malay
s
ia in
2012. He ser
v
ed as
a lectur
er
at Univ
ersiti
T
e
kn
ikal
Mal
a
ysia Mel
a
ka
(U
T
e
M).
Curren
t
l
y
, h
e
is a
PhD student
at
Univ
e
r
siti
T
e
kno
logi
Mala
y
s
ia
. His
research
int
e
rest
s are
in e
l
e
c
tri
c
al insul
a
tion
,
po
wer equipm
ents
& insul
a
tio
n
condition
m
onit
o
ring,
and r
e
new
a
ble
en
er
g
y
.
Imr
a
n Sutan C
h
air
u
l
receiv
e
d
the BSc. (Hons) in Electr
ical
Engineering (
I
n
dustrial Power)
from
Universiti
T
e
kn
ikal
Mal
a
y
s
ia Melaka (UT
e
M)
in 2008. In
2012, he receiv
e
d the Master
of
Ele
c
tri
cal
Engin
eering from
Uni
v
ers
iti
T
e
n
a
ga N
a
s
i
onal
.
Current
l
y
, he is
a
le
ctur
er at Univ
ers
iti
T
e
kn
ikal
Ma
la
ys
ia
Melak
a
(UT
e
M).
His r
e
sear
c
h
int
e
rests
in
clu
d
e
liquid
and
g
a
s insulat
i
on
as
well
as cond
itio
n m
onitoring of
power transform
e
r
.
Evaluation Warning : The document was created with Spire.PDF for Python.