HAL Id: hal-00141591
https://hal.archives-ouvertes.fr/hal-00141591
Submitted on 20 Apr 2007
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Static electrication of pressboard/oil interface and
transient phenomena
L. Peyraque, Abderrahmane Beroual, François Buret
To cite this version:
L. Peyraque, Abderrahmane Beroual, François Buret. Static electrication of pressboard/oil interface
and transient phenomena. IEEE Transactions on Dielectrics and Electrical Insulation, Institute of
Electrical and Electronics Engineers, 1998, 5 (3), pp.443-449. �hal-00141591�
IEEE
Transactions
01
Dielectrics and Electrical Insulation
Vol.
5
No.
3,
June
1998
t
v
Static Electrification of Pressboard/Oil
Interface and Transient Phenomena
ABSTRACT
The static electrification phenomenon of insulating materials used
in
power transformers is in-
stigated through
two
devices. The first is a cell with a rotating disk covered
on
both sides
L.
Peyraque
Jeumont Schneider Transformateurs, Lyon
M
d
t
g
a
e
a
a
II
r'
A.
Beroual
and
F.
Buret
Ecole Centrale de Lyon,
CEGELY
Ecully France
ith a given pressboard and immersed in a metallic tank containing
-4
1 of oil; and second
vice enables us to measure the electrostatic charge tendency of oils. The electrostatic charge
ndency
(ECT)
of insulating oils and the leakage current generated by the charge concentration
adient at the oillpressboard interface are analyzed as function
of
temperature, water content
d aging
of
oils, the nature of the pressboard, and the rotating speed of the disk. The influ-
ce of the surface roughness
of
the pressboard and antistatic additives
on
the aging and
ECT
so
are considered.
A
correlation between
ECT
and the physico-chemical and electrical char-
teristics
of
the oil is established. It
is
shown that the temperature gradients and the water
igration phenomena at the oillpressboard interface play an important role in the charge sepa-
tion. Transient currents are strongly affected by the presence
of
air
in
the pressboard and the
443
MOTOR
-
is in relation with the
s
failures of power transformers, re-
ing of the static electrifica nomenon. Indeed, the latter depends
ly on the behavior
of
these materials,
surements
of
the curre
lating materials on
ECT
of
the oil
is
attempted.
4
PRESSBOARDS
METALLIC
DISK
Figure
1.
Rotating disk
cell.
2
EXPERIMENTAL TECHNIQUES
Two experimental arrangements are used. The first one is a cell with a
rotating disk, known as
CIGRE
test cell allowing
us
to measure the cur-
rent induced
by
the charges created at the contact
of
insulating press-
boards and oils. The second arrangement enables
us
to characterize the
charging tendency
of
oils.
2.1
ROTATING
DISK
CELL
This resulted from experimental techniques used by petroleum in-
dustries and those used in the past by electrotechnic industries
[5].
It
1070-9878/98/
$3.00
0
1998
IEEE
444
Peyraque
et
al.:
Static Electrification
of
Pressboard/Oil
FARADAY
CAGE
VALVE
-
HI
I
k-t-
SYRINGE
OIL
I-
I1
1-1
I
FILTER HOLDER
TO ELECTROMETER
METALLIC
VESSEL
Figure
2
Charging tendency measuring apparatus
0
100
200
300
400
500
R
otatlon
speed
(roundlmln
)
Figure
3
Comparison
of
insulating materials in aged
oil.
250
2w
0
0
100
2W
303
4w
500
R
otaiion
speed
(roundfrnln,)
Figure
4.
Comparison
of
insulating material in new oil.
consists of a disk of
150
mm diameter and
6
mm thickness, covered
on
both faces with pressboard and immersed in
a
metallic tank of
250
mm
diameter and
80
mm
height (Figure
1).
The disk is placed at the center of
the tank where it can be rotated by
a
small electrical motor. The tank
as
well
as
the rotating disk are made
of
aluminum.
A
heating system allow-
ing temperatures between
20
and
90°C
is used. This apparatus allows
us
to quantify the charge created at the pressboard/oil interface through
a
current measurement. Indeed, thanks to an electrometer inserted be-
tween the disk axis and the ground, we measure the current generated
by the charge concentration gradient created at the pressboard/oil in-
terface. This current
IL
called ’leakage current’ is measured during
a
period
of
30
min; the transient current appearing just after the start of
the disk rotation
(i.e.,
the first
10
min
of
the test) is not taken into account.
The
CIGRE
cell and the electrometer are placed in
a
Faraday cage.
4M)
400
-350
5z
5
150
1CO
50
0
Roughness
(IrmI
-.-
Nswoil
-0-
&ad
oil
Figure
5.
Influence
of
the roughness
of
the pressboard on electrification
current with new and
aged
oils.
The rotation speed is
430
rpm.
W
bg
5;
U50
540
ka
0
0
02
0,4
0,8
48
1
(4
a,
CLm
G-m
540
v
-.=
tm
5:
0
5
10
15
Time
(h
90
“a,
wm
270
S%
ta
7aD
O
101
I
I
I
I
I
I
0
50100150”JD
(4
Time
(h
>
Figure
6.
Continuous measurement during
1
h
(a);
15
h
(b);
260
h
(c).
The rotation speed was
130
rpm.
2.2
CHARGING TENDENCY
APPARATUS
To complete the above measurement, we use another arrangement
allowing us to measure the
ECT
of a
given oil when it passes through
a
cellulosic filter
[6]
(Figure
2).
This allows us to compare the different oils
and to correlate their charging tendency to their dielectric and chemical
IEEE
Transactions
OII
Dielectrics and Electrical Insulation
Vol.
5
No.
3,
June
1998
445
1400
I
1200
%
-
1000
E
800
2
5
600
400
200
0
properties. Its working
neutral, through the filter,
terface. The oil is positively
surement is carried out by
rounding the filter. This
man
541
type), normally
L
at each new measurement,
1600
pIinciple consists of forcing an oil, electrically
and results in a charge separation at the in-
charged and the filter negatively. The mea-
an electrometer on the metallic support sur-
filter is a highly porous cellulose sheet (What-
sed to filter particles
of
3
pm.
It is changed
,
0
2c
Figure
7.
Effect
of
the te
ent combinations of pressbcards
of
130
rpm.
3
RESUI~TS
AND
DISCUSSION
40
60
ao
Temperature
(OC)
nperature rise on the leakge current for differ-
(C1
and
C2)
and oils, with
a
rotating speed
30%
Figure
8.
Effect
of
with a rotating speed of
13C
Table
1.
Characteristics of oils used in the tests.
m
mc/m3
0.2~10-*
New
oil
(Hlj
Aged
oil
(H2j
66
2.8X10-4
tempvrature gradient on pressboard
C1
and
new
oil
rpm.
TI~*.~N-
DOSE
10
'C
lo00
0
-
n
a
*
amo
2
am0
3
4ma
4
-1mo
W
am
am0
0
50
100
t50
200
?SO
300
Time
(h)
Figure
9.
Effect of temperature gradient on pressboard
C2
and aged oil
with
a
rotating speed of
130
rpm.
3.1
INSL
To compare the differenl
transformers, four types
of
in transformer) with differmt
35,50
and
100
pm
corresponding
is a synthetic material
few charges compared to
The tests are achieved
naphtenic oil: a new oil
(H:
these oils are given in Table
two times. At each test, dif
It appears from the
the leakage current varies
tween
80
and
430
rpm (FigL
IWO
-
4
,
c-i------f
-
I
I
i
INFLUENCE OF THE
LATING PRESSBOARD
pressboards which could be used in power
2
mm thickness (a common thickness used
surface roughnesses are considered:
15,
respectively to
C4,
C1, C3
and
C2. C4
(consisting of aramid NomexTM fibers); it creates
cdlulose materials.
in the
CERE
cell, filling with two kinds
of
)
and an aged oil
(H2);
the characteristics
of
1.
With the aged oil, the tests are repeated
:erent
oils
and pressboards are used.
experimental results that, for a given pressboard,
quasi-linearly with the disk rotation speed be-
res
3
and
4)
confirming the results reported
I
I
I
10
15
20
25
30
WATER CONTENT
(p p
m
)
Figure
10.
Influence
of
moisture
on
ECT
of
mineral
insulating oils.
elswhere by
us
[7]
and others
[5,8].
This linear increase of the current
has been attributed to the increase of the radial component
of
the circu-
lating velocity
of
oil on the disk. Such a linearity has been observed also
when studying laminar flow
[9,
lo].
On the other hand, the pressboard
C2
whose the surface roughness is
-3x
that
of
C1
creates a current
2.5
to
4x
higher than
C1; C3
which has a surface roughness
1.5~
higher
than
C1
generates in its turn a current only
1.5
to
2x
higher than
C1.
This linear variation tends to demonstrate that the smoothness of
C4
is
responsible for the very low current observed with this material. Note
that similar variations have been obtained on the materials
C1, C2
and
C4
on a tubular experimental arrangement
[lo,
111.
We also note a strong difference between the solid insulating mate-
rials with both new and aged oils. For given pressboard and rotating
velocity
of
the disk, the aged oil gives a leakage current
~2x
that
of
vir-
gin oil. Figure
5
gives the measured current
os.
the surface roughness
of
pressboards in both oils, for a rotating speed of the disk of
430
rpm. One
can note that the variation
of
the current is also quasi-linear, indicating
the importance of the surface roughness of the pressboard.
10 15 20
25
30
OIL
WATER
CONTENT(p.p.m
)
Figure
11,
Leakage current
vs.
water content of oils with
a
rotating
speed
of
130
rpm.
8o
T
70
1
E
504
-
LEAKAGE CURRENT
si\\h,I'
-+
0-
:
,
-.
0
50
100
150
200
250
300
Time
(hl
Figure
12.
Transient current
for
a
dried pressboard immersed
in
a
wet-
ted oil with
a
rotating speed of
130
rpm.
3.2
TEMPERATURE GRADIENT AND
TRANSIENT EFFECTS
3.2.1 STEADY STATE AT
ROOM
TEMP
E
RATU RE
By
carrying tests at room temperature during
a
period of 260 h, with
a
constant rotation speed of 130 rpm, we observe three phases in the
evolution of the leakage current (Figure 6). During the first 10 min, an
important peak appears on the leakage current.
After this peak, the current increases slowly during
a
longer period
(3
to
4
h).
This period corresponds to the thermal time constant
of
the
insulating oil. Thus the slow increase
of
the current probably is due to
internal heating induced by the movement of the disk. From
4
h after
the beginning
of
the test to the end
of
the test (260 h), the current is stable
at a fixed value.
160
T
P
-
1
1
LEAKAGE CURRENT
E
120
4
:-a
100
-PE
"U
p5
::I
l\
WATER CONTENT
$g
40
I
20
0
I
/
Y
0
50
100 150
200
250
300
Time
(h)
Figure
13.
Transient current for
a
wetted pressboard immersed in
a
dried
oil
with
a
rotating speed
of
130
rpm.
Peyraque
et
al.:
Static Electrification
of
Pressboard/Oil
1000
100
z
3
I-
V
10
1
1E-14 1E-13 1 E-1
2
1 E-1 1 1E-10
CONDUCTIVITY
lS/m)
Figure
14.
ECT
of
mineral oils vvs. the conductivity at ambient.
3.2.2 THE INTERFACE AND
With
a
new oil, the leakage current also increases with the tempera-
ture as reported previously
[7].
Figure
7
shows the influence
of
the tem-
perature rise on the leakage current for different combinations
of
press-
boards (C1 and
C2)
and oils (H1 and
H2),
in the range 20
to
80T,
on a
short period
of
4
h with a rotating test cell of 130 rpm.
Now, if we apply
a
longer heating period at
90T,
with nitrogen gas in
the upper tank of the cell
to
avoid
a
degradation of the oil, the behavior
will be different. Figure
8
gives the results obtained with the new oil and
pressboard
C1.
The same tests had been carried out with pressboard
C2,
and a slightly aged oil (Figure
9).
From these tests, we can deduce that the leakage currents consists
of
three phases following a rise in the oil temperature: a strong rise of the
current over a short period
of
10
h;
a
slow decrease during
60
to
80
h
and
a
stable value higher than that observed at room temperature.
TEMPERATURE GRADIENT
Oil
B
2
1000
I-
-
Oil
C
S
Oil
D
100