December 2010, Volume 4, No.12 (Serial No.37)
Journal of Energy and Power Engineering, ISSN 1934-8975, USA
Insulation System of Power Transformers—Behavior
Comparison of Mineral Oils and Natural Ester Dielectric
Fluids
R. Polanský, P. Prosr, V. Mentlík, J. Pihera and P. Trnka
Department of Technologies and Measurement, Faculty of Electrical Engineering, University of West Bohemia, Univerzitní 26, 306
14 Pilsen, Czech Republic
Received: June 19, 2010 / Accepted: September 01, 2010 / Published: December 31, 2010.
Abstract: Paper deals with a comparison of selected properties of several vegetable oil representatives along their accelerated thermal
ageing at the temperature of 90 ℃. These properties are compared to two widely used and commercially available mineral transformer
oils. A combined insulating system (an oil-paper system) was created with the usage of mentioned oils for measurement purposes.
Dissipation factor, capacity and volume resistance are characteristics measured along a thermal ageing of the oil-paper systems.
Infrared spectroscopy was used as an additional method. After 1,000 hours of ageing, the dissipation factor of all systems based on
vegetable oils did not exceed the value of 0.015. The volume resistance of systems containing mineral oils was approx. twice as high as
the volume resistance of those with vegetable oils. The capacity on the other hand was slightly lower in the case of mineral oils
application. An experiment also showed that the paper combined with the vegetable oil dries more quickly than in combination with the
mineral oil. Infrared spectroscopy has not shown any expressive changes in the chemical structure of all tested oils yet (up to 1,000
hours of ageing).
Key words: Power transformer, ecosystem, biodegradable oils, electrical properties.
1. Introduction
During a power transformer operation, some critical
situations (such as an atmospheric discharge, a
turn-to-turn fault, a diverter switch failure, etc.), can
occur. Such situation could lead to the fire of the
transformer, which would threaten both employees and
environment. Soil contamination caused by the
transformer oil leakage or air contamination caused by
burning products could also occur as a side effect. These
situations have to be considered despite a low
probability of their occurrence. To avoid environmental
disaster we (besides the design set-ups, e.g., a flood pool)
Corresponding author: R. Polanský, Ph.D., associate
professor, research fields: dielectrics and electrical insulation
(diagnostics and development of new electrical insulating
materials for electrical engineering, analytical chemistry,
application of thermal analyses, study of aging process). E-mail:
rpolansk@ket.zcu.cz.
look for an acceptable replacement of currently used
mineral (petroleum based) transformer oils. The mineral
oils are known for their high flammability and very bad
biodegradability [1] on the one hand and excellent
properties (electrical, chemical, etc.) on the other hand.
Transformer oils of the “new era” should have very
good biodegradability, better thermal resistivity, stable
electrical properties and also reasonable price.
Therefore the main attention is currently given among
others to financially reasonable and easily available
vegetable oils (natural ester dielectric fluid) [2].
This problem has been solved since 1980’s and more
and more attention is paid on it today [3-8]. It is known
about the natural esters that they have greater affinity
for water than do the mineral oils, so less water will
remain in the paper as the water divides itself between
headspace, the paper, and the fluid. Water is consumed
52
Insulation System of Power Transformers—Behavior Comparison of Mineral Oils and Natural Ester
Dielectric Fluids
by hydrolysis of the natural ester, producing free fatty
acids. These may react with a cellulose backbone via a
transesterification and protect the cellulose from
hydrolysis [4, 6]. There is rich experience with
practical application of these oils, e.g., McShane deals
with these problems with great intensity. As written in
his paper [3], the first prototyped natural ester-filled
transformers were installed in 1996. Since then, more
than 150,000 distributions have been installed. More
than 200 substation transformers have been built or
retrofilled with the natural ester fluid currently up to
200 MVA in base ratings and at voltages as high as 242
kV. Luksich also stated in his presentation at
Weidmann-ACTI 3rd Annual Technical Conference in
Sacramento in 2004 that several transformers filled
with the natural ester fluids have been accessible for a
periodic fluid testing, including dissolved gas analysis.
Changes in these transformers over time correspond to
changes predicted by an accelerated life testing. In
short, the types of the changes seen in the mineral oil
are also typically seen in the natural ester fluids [7].
Another experience with a practical application of the
natural esters is described, e.g., in the work of Stockton
[5]. Presented statements refer to the promising future
of these perspective fluids.
2. Objectives
The main objective is to compare electrical
properties and infrared spectra of chosen vegetable oils
(the oil-paper systems respectively). Commercially
available ENVIROTEMP® FR3TM vegetable-based
transformer oil [8], sunflower oil, rape oil and refined
sunflower and rape oil were exposed to the accelerated
thermal ageing at the temperature of 90 ℃ for times
ranging from 0 to 1,000 hours. These oils were at the
same time compared to two widely used and
commercially available mineral transformer oils
(Technol Y 3000 oil and Shell Diala DX oil).
system) was created by diving of flat paper specimens
(size of 100×100 mm, thickness of 1 mm) to the tested
oils. During the whole experiment, the oils together
with the paper specimens were placed in glass vessels
and stayed closed along the thermal ageing by reason
of partial restriction of air access.
2.2 Method Description
A standard three-electrode system was used for the
measurement of electrical characteristics (dissipation
factor tg δ [-], capacity Cx [F] and volume resistance Rx
[Ω]). The electrode system was dived in the tested oil
together with the paper specimen every time, when
sampling was done-see Fig. 1.
During the dissipation factor and capacity
measurement, the three-electrode system was
connected to an automatic Schering Bridge (LDV-5
Lemke Diagnostics GmbH), during the volume
resistance measurement it was connected to an
electrometer with an internal source of DC voltage of
500 V (Keithley Instruments, model 6517).
Infrared spectra (FT-IR spectra) were recorded on
pure oils via Attenuated Total Reflectance technique
(ATR) on a Nicolet 380 spectrometer (Nicolet
Instrument). The spectrometer was being purged by dry
air during the measurement. The measurement was
proceeded with a ZnSe crystal, when 32 scans with the
resolution of 4 cm-1 were collected for each specimen.
2.1 Specimen Preparation
The combined insulating system (the oil-paper
Fig. 1 Electrical test set-up.
Insulation System of Power Transformers—Behavior Comparison of Mineral Oils and Natural Ester
Dielectric Fluids
53
Dissipation factor tan δ [-]
0,14
0,12
0,1
0,08
0,06
0,04
0,02
0
0
50
125
225
500
1000
Time of thermal ageing [hours]
Shell Diala DX
Sunflower
Envirotemp FR3
Sunflower oil - refined
Rape oil
Technol Y 3000
Rape oil - refined
Fig. 2 Thermal ageing.
+0
50
125
225
500
T ime of thermal ageing [hours]
0E
0
2,
2.3 Thermal Ageing
8
2,
2E
The subsequent analysis of the measured spectra was
performed by an OMNIC software.
V o lu m e re s is ta n c e R x [ Ω ]
4,
6,
8,
1,
1,
2
2
0
2
2
+0 E +0 E +0 E +0 E +1 E +1
9
9
9
9
0
0
Fig. 3 Dissipation factor vs. thermal ageing.
Shell Diala DX
Envirotemp FR3
Rape oil
Sunflower
Sunflower oil - refined
Technol Y 3000
Fig. 4 Volume resistance vs. thermal ageing.
filter paper (8 μm thickness of screen hole). This
refining enabled to remove the free fatty acids.
3. Results
3.1 Electrical properties
Following graphs (Figs. 3-5) show the results of the
electrical properties measurements.
The dissipation factor dependence on the time of the
thermal ageing is presented in Fig. 3. Although all oils
and the tested paper had been dried before testing (at
6,
0E
-1
1
RCOOH + NaOH → RCOONa + H 2O (1)
0.6 mg of NaOH was added to 1 g of the oil. Sediments
(coagulations) were removed through the usage of a
Rape oil - refined
Capacity C x [F]
1,
8,
0E
0E
-1
-1
1
0
Besides the measurement of above mentioned
properties on virgin oils, these properties were
measured also along the accelerated thermal ageing at
90 ℃ under the limited access of air (see Fig. 2).
The sampling was realized at the times after 0, 50,
125, 225, 500 and 1,000 hours of the thermal ageing.
The sunflower and rape oils were also measured
after an alkaline refining according to Li [9]:
1000
0
50
125
225
500
Time of thermal ageing [hours]
Shell Diala DX
Envirotemp FR3
Rape oil
Sunflower
Sunflower oil - refined
Technol Y 3000
1000
Rape oil - refined
Fig. 5 Capacity vs. thermal ageing.
60 ℃ for 48 hours), high levels of the dissipation factor
were recorded at first (tg δ = 0.1-0.124, i.e., 10-12.4%
of dielectric losses), which is most likely caused by
high portion of residual moisture.
This value improves expressively with the
increasing time of thermal aging and the dissipation
factor thus refers to the proceeding drying of the paper
very well. The fact, that the vegetable oils reach the
optimum levels of the dissipation factor (approx.
Insulation System of Power Transformers—Behavior Comparison of Mineral Oils and Natural Ester
Dielectric Fluids
3.2 Fourier Transform Infrared Spectroscopy
An FT-IR spectroscopy was measured on pure oils
removed from the vessels. The results of the FT-IR
spectroscopy refer to differences between the vegetable
and mineral oils very well from the view of their
chemical structure. Following Fig. 6 compares these
differences of particular oil specimens.
0,10 Rape seed oil - virgin state
Ab s
interval of 500-1,000 hours (see cut in Fig. 3), tan δ
level of all tested insulating systems stabled in the
range from 0.005 (Technol Y 3000) to 0.015
(Sunflower oil - refined).
The moisture in the inner structure of the tested
specimens was also proven by the volume resistance
evolution in dependence on the time of thermal ageing.
(Fig. 4).
As obvious, the expressive increase of the volume
resistance of all tested specimens was observed over
the time owing to the thermal ageing (a 2-order
increase was observed). Also in this case, the paper
combined with the vegetable oils dries more quickly up
to 225 hours of thermal ageing than does with the
mineral oils. This tendency changes in the time interval
from 225 to 1,000 hours.
Further the capacity influence on the time of thermal
ageing is presented in Fig. 5.
The capacity level is closely associated with the
level of relative permittivity. The higher the capacity (~
the relative permittivity) is, the better is the ability of
the dielectric to accumulate the electric charge, which
is much undesired in the field of power transformers.
Compared to the previous plots, the difference between
the vegetable and mineral oil groups is obvious from
the very beginning of capacity measurement. The
insulating systems with the mineral oils have slightly
lower capacity level in general than the systems
containing the vegetable oils.
The most significant difference in the oil spectra was
observed in a presence of the strong band of 1,743 cm-1
at the vegetable oils (a carbonyl stretching vibration),
which marks an oxidative stability.
In general, all oils containing oleic, linoleic and
linolenic acid have lower oxidative stability
considering a higher concentration of polyene fatty
acids [10].
Another difference can be observed in a fingerprint
region from 1,000 to 1,300 cm-1. It is one of the specific
spectrum regions, where vibration symmetry is very
complicated and the final absorption spectrum is
unique both for whole molecules or their functional
groups.
Meanwhile, no significant change of obtained spectra
of tested oils was observed when comparing the oil
spectra along the thermal ageing. In the mineral oils,
(see spectra of Technol Y 3000 oil in Fig. 7) an
intensity has not changed neither in the spectral bands
of 3,000 and 1,460 cm-1, which correspond to
hydrocarbons, nor in the bands in the region from 800
to 712 cm-1, corresponding to the refining level.
Only the slight change of intensity was observed in
the frequency of 1,157 cm-1.
The thermal ageing influenced the vegetable oils
spectra even less than the mineral oils spectra. This is
obvious in Fig. 8, which shows the sunflower oil
spectra for an example. As evident, the single spectra
almost overlap each other.
0,05
0,00
0,10 FR3 - virgin state
Ab s
0.05-0.01) more quickly than the mineral oils, is very
interesting (and it is fully in accordance with the
conclusions of other authors [6]). This tendency was
observed up to 500 hours of thermal ageing. In the
0,05
0,00
0,10 Diala - virgin state
Ab s
54
0,05
0,00
4000
3000
2000
Wavenumbers (cm-1)
Fig. 6 Comparison of mineral and vegetable oils.
1000
Insulation System of Power Transformers—Behavior Comparison of Mineral Oils and Natural Ester
Dielectric Fluids
Technol - virgin state
changes in the chemical structure of the tested oils yet.
0,08 Technol - 225 hours
Technol - 1000 hours
Absorbance
0,06
0,04
0,02
0,00
3000
2000
Wavenumbers (cm-1)
1000
Fig. 7 IR spectra of mineral oil vs. thermal ageing (Technol
Y 3000).
0,10 Sunflower - virgin state
Sunflower - 225 hours
Sunflower - 1000 hours
0,08
Absorbance
55
0,06
0,04
• All observed electrical properties (Rx, tan δ and
Cx) of the paper combined with the commercially
available ENVIROTEMP® FR3TM vegetable-based
transformer oil are practically the same as in
combination with the rape oil. Also the IR spectra of
these two oils (Fig. 6) are almost identical.
• The detailed analysis indicated the properties of
the paper and rape oil combination as the best from all
tested vegetable specimens.
All obtained results confirm the good properties of
the mineral oils. This fact is caused mainly by their
chemical structure. As mentioned above, big accent is
put on the application of environmentally friendly
materials nowadays, which requires as intensive
research of the alternative insulating oils (vegetable,
synthetic oils, etc. [11]) as that done in the past with the
mineral oils.
Acknowledgments
0,02
Fig. 8 IR spectra of vegetable oil vs. thermal ageing
(sunflower oil).
This article was carried out with the support of
Ministry of Education, Youth and Sports of Czech
Republic,
MSM
4977751310–Diagnostics
of
Interactive Processes in Electrical Engineering.
4. Conclusions
References
Following conclusions after 1,000 hours of thermal
ageing can be drawn:
• The value of dissipation factor of all insulating
systems containing the vegetable oils does not exceed
the value of 0.015. In spite of surprisingly good result,
this value is still approx. 2-3 times higher than the
value of the mineral oils.
• The volume resistance of systems containing the
mineral oils is approx. twice as high as the volume
resistance of systems with the vegetable oils.
• The capacity is slightly lower in the case of mineral
oils application then the capacity of vegetable oils.
• The paper combined with the vegetable oil dries
more quickly than in combination with the mineral oil.
• The IR spectroscopy has not recorded any expressive
[1] I. Fofana, A. Bouaïcha1, M. Farzaneh, J. Sabau, Ageing
behaviour of mineral oil and ester liquids: a comparative
study, in: IEEE Electrical Insulation and Dielectric
Phenomena, Québec City, Canada, Oct. 26-29, 2008, pp.
87-90.
[2] C.P. McShane, G.A. Gauger, J. Luksich, Fire resistant
natural ester dielectric fluid and novel insulation system for
its use, in: IEEE/PES Transmission & Distribution
Conference, 1999, pp. 890-894.
[3] C.P. McShane, Natural ester dielectric fluid development
update, in: IEEE/PES Power & Energy Society General
Meeting, July 26-30, 2009, pp. 1-6.
[4] C.P. McShane, J.L. Corkran, K.J. Rapp, J. Luksich, Aging
of paper insulation retrofilled with natural ester dielectric
fluid, in: IEEE/DEIS Conference on Electrical Insulation
and Dielectric Phenomena, Oct. 19-22, 2003, pp.124-127.
[5] D.P. Stockton, J.R. Bland, T. McClanahan, J. Wilson, D.L.
Harris, P. McShane, Natural ester transformer fluids: safety,
reliability & environmental performance, in: IEEE/PCIC
0,00
4000
3000
2000
Wavenumbers (cm-1)
1000
56
Insulation System of Power Transformers—Behavior Comparison of Mineral Oils and Natural Ester
Dielectric Fluids
Petroleum and Chemical Industry Technical Conference,
Sept. 17-19, 2007, pp. 1-7.
[6] J.Y. Li, J.L. Rui, X.S. Cai, G.S. Hui, Study on the
influence of natural ester on thermal ageing characteristics
of oil-paper in power transformer, in: IEEE/ICHVE
International Conference on High Voltage Engineering and
Application, Chongqing, China, Nov. 9-13, 2008, pp.
437-440.
[7] J. Luksich, Evaluating new and in-service vegetable oil
dielectric fluids, presented at the Weidmann-ACTI 3rd
Annual Technical Conference, Sacramento, CA, Nov. 8-10,
2004.
[8] Envirotemp® FR3TM Fluid, Bulletin 00092, Cooper
Power Systems, Inc., 2005.
[9] X.H. Li, J. Li, C.X. Sun, Properties of transgenic rapeseed
oil based dielectric liquid, in: IEEE SouthEastCon,
Memphis, TN, 2006, pp. 81-84.
[10] P. Kadlec, et al., Food Technologies II (in Czech), The
Institute of Chemical Technology, Prague, 2002, pp. 89.
[11] V. Mentlík, R. Polanský, P. Prosr, J. Pihera, P. Trnka,
Synthetic ester-based oils and their application in power
industry, in: International Conference on Renewable
Energies and Power Quality, Valencia, Spain, April 15-17,
2009.