Ester Fluids As Alternative for Mineral Based

Transformer Oil
H. Boris, E. Gockenbach, B. Dolata
Institute of Electric Power Systems
High Voltage Engineering Section, Schering-Institute
Leibniz Universität Hannover
Callinstr. 25A, 30167 Germany
dolata@si.uni-hannover.de

Abstract- One of the key component of electrical power
distribution networks is transformers, interconnecting between
transmission and distribution voltage level or between different
networks. Cooling affords a liquid immersed insulation for high
power use, thus the insulation of these devices usually consists of
mineral oil and cellulose based porous material immersed
therewith. Synthetic ester fluids as well as natural based ester
fluids will be presented in this article. The electrical and
dielectrical properties and the measurements procedures will be
described.
I. INTRODUCTION
As mineral insulating oils which are one of the key
component of the insulation comprise several risks concerning
aging, fire resistance and ecology, the use of ester liquids
become increasingly attractive. Moreover, the price of mineral
oil based products has been continuously escalating. In this
contribution new ester insulating liquids are introduced that
combine the advantages of mineral insulating oils and
ecological friendly, low viscosity synthetic ester liquids.
Insulating liquids of mostly synthetic origin like High
Molecular Weight Hydrocarbons or Silicone Oils could not so
far substitute mineral oils as their application characteristics
are inferior or their price is disproportionate. In Europe only
synthetic ester liquids, named Midel 7131, have proven their
advantageousness for transformers regarding environmental
and aging aspects.
The developed ester liquids meet the major and important
conditions for new insulating liquids that are environmental
friendliness, high fire resistance and lower viscosity.
Therefore, they are very promising liquids for use as cooling
and insulation liquid in transformers and another high voltage
apparatuses.
This contribution presents some experiments on the
electrical and dielectric characteristics of these liquids, which
have been performed to approve ther usability for
transformers.
Mainly the Breakdown Voltage (UB), the Dielectric
Dissipation Factor (tanδ), the Relative Permittivity (εr) and the
Partial Discharge Inception Voltage (UPDIV) characterize an
insulating liquid and are required for a qualified dimensioning
of a transformer.
II. MINERAL INSULATING OIL
Because of the wide availability and the cost benefit that
result from this kinship, petroleum-based transformer oils are
(probably) the most widely used electrical insulating liquids in
the world today - and have been for the past century [1].
Indeed, for most of high voltage transformers, the main
insulation is a combination of mineral oil with cellulose
materials.
Mineral insulating oils consist of different hydrocarbons,
namely Paraffines, Olefines, Naphtenes and Aromates. These
hydrocarbons are constituted of carbon and hydrogen linked
with mono and double bonds. Depending on the kind of
bondage the aging stability of the molecule is also
inconsistent. Due to stresses like pyrolysis (temperature),
hydrolysis (water) and the electrical field the bonds are prone
to cracking and therefore leading to different dissolved gases
that are used for Dissolved Gas Analysis (DGA). Catalysis
from the metals in the transformer additionally facilitates this
process.
Although this the use of mineral oils in power transformers
has proven to be very reliable it also comprises several
disadvantages. Besides the aging also the environmental
endangerment resulting out of a transformer spillage has to be
kept in mind.
III. ESTER FLUIDS
An alternative to the mineral oils are ester liquids consisting
of naturally or synthetically arranged constituents. Basically
esters can be classified in five groups:
• Monoesters
• Dicarboxylic esters (Diesters)
• Glycerinesters
• Polyolesters
• Complexesters
All kinds of esters have to be synthesized out of an acid and
an alcohol. Only Polyol- and Complexesters are adequate for
high stress conditions that it is required for use in a high
voltage transformer. These ester groups are the strongest
basestock among synthetic oils. The competence is originating
from the absence of secondary hydroxyl groups and a
quarternary carbon in the β - position. Typical alcohols used voltage rises of 1 kV/s up to the occurrence of stationary
for synthesis are neopentylglycol, trimethylol- propane, partial discharges with qs≥100 pC. The test vessels are
pentaerythritol or dipentaerythritol. The viscosity of the depicted in Fig. 2.
product can mainly be controlled in the group of
Complexesters which are polyols esterified with a blend of
mono-, bi- or tri-functional carbonic acids to form oligomer
molecules with a selectable sheer stability [3].
The new liquids have been designed for power transformer
use out of this “construction kit” and are the center of a whole
family of esters fitted for every kind of transformer application
(distribution, network, traction and machine transformers).

TABLE I
TECHNICAL DATA FOR ESTER FLUIDS
Midel 7131 s-Ester n-Ester 1 n-Ester 2
tan δ (%) at 20°C 0,001 0,0006 0,001 0,003
tan δ (%) at 90°C <0,03 <0,03 0,009 0,005
εr at 20°C 3,3 3,0 3,0 3,0 Fig. 2. The test vessel for the determination of UPDIV according to IEC 61294
Breakdown voltage
<75 99 <75 56
(kV)
Fire point (°C) 322 303 356 360 Dielectric Dissipation Factor tanδ and Relative Permittivity εr
Pour point (°C) -60 -50 -31 -21
Kinematical The Dielectric Dissipation Factor tan δ and the Relative
viscosity (mm2/s) at 28 25 37 33
40°C Permittivity εr are measured with a CPC 100 Multifunction
Primary Test System and a CP TD1 Tangent Delta system
Table I shows the technical specifications of producer. The manufactured by OMICRON. For these investigations the
letter s- symbolizes synthetic and n- natural ester fluid. temperatures 20°C and 90°C have been used. The electrodes
arrangements are depicted in Fig. 3.
IV. MEASUREMENT PROCEDURES

Breakdown Voltage
For the measurement of the Breakdown Voltage versus
temperature according to IEC 60156 [4] a test vessel has been
used as depicted by Fig. 1.

Verdrängungskörper Verdrängungskörper

Pt 100

Fig. 1. The test vessel for breakdown tests according to IEC 60156

The temperature have been chosen to simulate ambient 20°C, Fig. 3. The components of the test vessel and electrode arrangement for the
dielectric measurements (tanδ and εr)
operating 50°C as well as the temperature at overload 90°C
and Hot Spot temperature 110°C. The cylindrical shell around
As known from theory and scientific experience, the
the electrodes allowed a heating of the test setup up to 110°C
repeatability and accuracy of measurements play a very
by use of a thermostatic heater. For the measurement a rise of
important role to reach comparability of the results. For this
the voltage of 2 kV/s and the electrode gap of 2,5 mm has
reason, the same measurement condition and handling of
been chosen.
investigated samples was also necessary.
In order to meet the conditions for each investigated samples
Partial Discharge Inception Voltage
the test vessels have been cleaned and the electrodes polished.
For the determination of the Partial Discharge Inception
For the electrically and dielectric measurements the
Voltage (UPDIV) according to IEC 61294 [5] a test vessel with
investigated ester fluids have been tested in delivered, not
a needle-sphere and needle-plate electrode arrangement has
prepared condition. The water content occurred 60 ppm-
been used. The measurements occurred at 20°C. The test
70 ppm.
 V. RESULTS The investigations for the determination of the Partial
Discharge Inception Voltage have been analyzed for two
A. Breakdown Voltage different electrode arrangements. The needle-sphere and
The Breakdown Voltage has been determined and depicted needle-plate have been used to simulate different operation
as average value in kV. The standard deviation from six conditions. Fig. 5 shows the Partial Discharge Inception
measurements has been calculated. The Fig. 4 shows the Voltage and the standard deviation in kV. The stationary
Breakdown Voltage in dependence on the temperature in the partial discharges with qs≥100 pC are occurred up to 22,5-
range from 20°C to 110°C. 27,5 kV for the needle-sphere electrode and up to 27,5-
32,5 kV for the needle-plate electrode arrangement. As a result
90
Midel 7131
it can be easily seen, that the natural based ester liquids have a
kV
80 s-Ester good and comparable characteristic to synthetic ester fluids.
n-Ester 1
70 n-Ester 2
C. Dielectric Dissipation Factor tanδ
60
Fig. 6 shows the results of Dielectric Dissipation Factor tanδ
Breakdown Voltage

50 in dependence on the temperature. The ambient 20°C and Hot

40
Spot 90°C temperature has been used. It can be seen that
dielectric losses of investigated ester liquids are comparable.
30
Only synthetic ester liquid named s-Ester shows particular
20 variations. As dielectric losses only marginally contribute to
10
transformer heating due to losses this aspect is of minor
importance for transformer operation.
0
20 50 90 110 °C
Temperature 13,5
% Midel 7131
12,0 s-Ester
Fig. 4. Breakdown Voltage for different ester liquids versus temperature n-Ester 1
1,8 n-Ester 2
Dielectric Dissipation Factor tanδ

1,6
For the synthetic based ester liquids the standard IEC
1,4
61099[2] is valid. For this kind of pentaerythrit esters the limit
1,2
of the Breakdown Voltage of 45kV is required [9]. On the
1,0
other hand the definition of IEC standard for natural based
0,8
ester liquids is still in progress. For all types of ester liquid is
0,6
to observe that the Breakdown Voltage is about 50-60 kV at
0,4
the operating temperature of 50°C as well as the Hot Spot
0,2
temperature of 110°C. The average value of all the
0,0
investigated ester liquid is higher as 45kV and accomplishes 20 90 °C
the standard IEC 61099. Temperature

B. Partial Discharge Inception Voltage Fig. 6. Dielectric Dissipation Factor tanδ for different ester liquids

45 D. Relative Permittivity εr
Midel 7131
kV
40 s-Ester
n-Ester 1 4,5
Partial Discharge Inception Voltage

35 n-Ester 2 Midel 7131

4,0 s-Ester
30 n-Ester 1
3,5 n-Ester 2
25
3,0
Relative Permittivity εr

20
2,5

15 2,0

10 1,5

5 1,0

0 0,5
Needle-Sphere Needle-Plate
Electrode arrangement 0,0
20 90 °C
Temperature
Fig. 5. Partial Discharge Inception Voltage (UPDIV) for different ester liquids at
needle-sphere and needle-plate electrode arrangement at 20°C
Fig. 7. Relative Permittivity εr for different ester liquids
When looking at the relative permittivity as it is depicted in REFERENCES
Fig. 7 it can be seen, that the measured values are comparable
[1] IEC 60156 (VDE 0370-5): „Insulating liquids – determination of the
between the ester liquids. The relative permittivity of about
breakdown voltage at power frequency – Test method“
3,0 is an advantage for the design of the insulation as the [2] IEC 61099 (VDE 0375-1): „Specifications for unused synthetic organic
homogeneity of the electrical field at the interface between esters for electrical purposes“
cellulose (εr=5,6) and ester fluids (εr=3,0). [3] Wasserberg, V., Borsi, H., Gockenbach, E., Fofana, I.: „Drying of
Liquid Immersed Solid Insulations Using a Hygroscopic Insulating
Liquid“, 12th International Symposium on High Voltage Engineering,
E. Water Saturation Point Bangalore, India, August 2001
The water saturation point based on exponentially [4] IEC 61294 (VDE 0380-5): „Insulating liquids - Determination of the
correlation has been calculated. Fig. 8 shows the different partial discharge inception voltage (PDIV) – Test procedure”
[5] IEEE C57-100: „IEEE Standard Test Procedure for Thermal Evaluation
saturation point were wo and H are oil specific coefficients.
of Liquid- Immersed Distribution and Power Transformers”
[6] Ahlemeyer, W.: „Der Einfluss gelöster Gase auf die Alterung eines Öl-
8000 Papier- Dielektrikums”, Ph.D. thesis, University of Hannover, Germany,
ppm Midel 7131
s-Ester 1981
7000 [7] IEC 61203 (VDE 0375-2): „Synthetic organic esters for electrical
n-Ester 1
n-Ester 2 purposes – Guide for maintenance of transformer esters in equipment“
6000 -H/T+273,15 [8] Beyer, M., Boeck, W., Möller, K., Zaengl, W, „Hochspannungstechnik–
W Saturation = W 0 exp
Water Saturation Point

Theoretische und praktische Grundlagen für die Anwendung“, Springer

5000 W 0, H - oil specific coefficients
Publishing, Berlin, Heidelberg, New York, London, Paris, Tokyo, 1986,
ISBN 3-540-16014-0 , 0-387-16014-0, pp.146-147.
4000 [9] V. Wasserberg, B. Dolata, H. Borsi, E. Gockenbach, H. Baehr,
„Ecological Friendly Alternative to Mineral Based Transformer
3000 Insulating Oils with Superior Properties”, in Proc. 2005 International
Symposium on High Voltage Engineering, pp. 313.
2000

1000

0 20 40 60 80 100 °C
Temperature

Fig. 8. Water saturation point for different ester fluids

VI. CONCLUSION
The presented experiments have shown that ester liquids are
an alternative to mineral insulating oils for use as insulating
and cooling system in transformers. As with continuously
increasing price of mineral based products also the naturally
based esters become more attractive.
The Breakdown Voltage about 50-60 kV fulfilled the
standard IEC 61099 requirements. The investigations of
Partial Discharge Inception Voltage have shown that the
natural based ester liquids have a good and comparable
characteristic to synthetic ester fluids.
The relative permittivity of all investigated esters is
comparable (3,0) and is an advantage in designing an
insulation. The high saturation point as the case may be the
high water uptake capabilities of ester liquids can bee used by
retrofilling old transformers.
Ongoing experiments concerning aging and gassing behaviour
(DGA) will examine all remaining aspects valuable for
operation of ester liquids in all kind of transformers. Further
investigations e.g. on the gassing behavior are necessary to
complete the evaluation of the new ester liquids. Additional
tests of the breakdown behavior at large gaps will show the
usability of the new ester liquids for high voltage applications
in power transformers.