770 Y. Du et al.

: Effect of Semiconductive Nanoparticles on Insulating Performances of Transformer Oil

Effect of Semiconductive Nanoparticles on Insulating

Performances of Transformer Oil

Yuefan Du1,3, Yuzhen Lv2, Chengrong Li1, 3, Mutian Chen1, 3, Yuxiang Zhong1,3,
Jianquan Zhou1,3, Xiaoxin Li2 and You Zhou1,3
1
Beijing Key Laboratory of High Voltage and EMC, School of Electric and Electronic Engineering, North China Electric Power
University, Beijing 102206, China
2
School of Energy, Power and Mechanical Engineering, North China Electric Power University, Beijing 102206, China
3
State key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, North China Electric Power
University, Beijing 102206, China

ABSTRACT
In this paper, TiO2 semiconductive nanoparticles with a large relaxation time constant is
added into transformer oil to form semiconductive nanofluids (SNFs), with the aim of
enhancing insulating characteristics. ac, dc and lightning impulse breakdown voltage
and partial discharge (PD) characteristics of oil samples before and after modification
were measured according to ASTM standard methods. It was found that SNFs have ac,
dc and lightning impulse breakdown voltage up to 1.2 times compared with pure oil.
Meanwhile, the partial discharge resistance of SNFs was also dramatically improved.
Charge trap and transportation characteristics of both samples have been measured by
thermally stimulated current method (TSC) and pulse electroacoustic technique (PEA).
It was found that electron shallow trap density and charge decay rate are greatly
increased in semiconductive nanoparticles modified transformer oil. It is proposed that
electron trapping and de-trapping processes in the shallow traps could be one of the
main charge transport processes in dielectric liquids.

Index Terms — Dielectric liquid, nanotechnology, semiconductor materials,

insulation, thermally stimulation current, acoustoelctric transducers, charge carrier
processes.
up to 1.1 times when the moisture content was 10 ppm.
1 INTRODUCTION Furthermore, the positive impulse breakdown voltage of
magnetic nanofluids was improved up to 2 times that of the
TRANSMISSION and distribution transformers form a base oil [4-5]. Based on simulating the propagation of streamer,
critical, highly loaded and expensive part of the electricity a model of electron scavenging by conductive nanoparticles is
generation and distribution network. Electric transformers rely proposed to explain the cause of the decrease in positive
on the high dielectric strength and cooling properties of streamer velocity and higher breakdown strength in magnetic
insulating oil to achieve normal operation [1]. The potential NFs stressed by positive impulse voltage [6-7]. It is believed
consequences of transformer failure can be quite damaging [2]; that the magnetite nanoparticles in NFs can capture free fast
and dielectric insulation problems blamed for 75% of high electrons required for streamer development and convert these
voltage transformer failures [3]. It is therefore important to fast electrons to slower negative charged particles, because the
improve insulating properties of transformer oil. relaxation time constant (τ=7.47×10-14 s) of the conductive
A particularly novel work aiming to enhance transformer magnetite nanoparticles is much faster than the microsecond
oil’s insulating characteristic is development of dielectric time scale of streamer development in transformer oil.
nanofluids (NFs) by suspending nanoparticles in transformer However, Kopwansky et al have found that the dispersibility
oil. Recently, much effort has been focused on preparing of magnetic nanoparticles is greatly influenced by orientation
magnetic NFs by suspending magnetite nanoparticles into the of external magnetic field. In magnetic field H||E the magnetic
transformer oil and testing their insulating and thermal property. particles aggregation led to the formation of the bridge across
It has been shown that ac breakdown strength can be increased the gap between the electrodes, which lowered the dielectric
breakdown voltage [8]. In addition, Chiesa et al [9] have
Manuscript received on 30 August 2011, in final form 21 March 2012. demonstrated that a kind of conductive nanoparticles (σ =1×102

1070-9878/12/$25.00 © 2012 IEEE

IEEE Transactions on Dielectrics and Electrical Insulation Vol. 19, No. 3; June 2012 771
-12
S/m) with a short relaxation time constant (τ =1.1×10 s) fails electrodes was chosen in the following tests. The initial
to increase the breakdown strength of transformer oil. The standing time was 5 minutes, the time interval between two
model of electron scavenging by conductive nanoparticle, successive shots was fixed at 1 minute. Apply three impulse
judged by the relaxation time constant fails to explain these waves at each voltage level and it is necessary to have at least
phenomena. Thus, there is still more work to do for exploring three withstand levels prior to breakdown. All experiments
the modification mechanism of nanoparticles on insulating were performed at room temperature. After one breakdown
properties of transformer oil. occurs, the oil sample and electrodes were changed, totally 10
In this paper, we aim to develop a new type of nanofluids by samples were carried out for each type of oils. In lightning
suspending semiconductive nanoparticles into transformer oil, breakdown tests, the liquid sample and electrode should be
investigate their insulating properties and propose the frequently changed due to high energy damage of impulse shots,
modification mechanism based on trap characteristics and which increases the difficulties of large sample size test when
charge transportation behaviors of nanofluids. considering the constraints of manpower, finance and so on. So
finding a proper way to carry out the impulse breakdown tests
is still of importance.
2 EXPERIMENTAL DETAILS
The high voltage dc generating circuit is capable of
2.1 Preparation of Nanofluids generating up to 200 kV. The voltage rising at a rate of 1 KV/s
was used to stress the samples. 10 mm gap of needle with a 3
The mineral oil (25# Karamay) was filtered through a μm radius of curvature to sphere electrodes was chosen in the
membrane filter in order to remove the impurity particles and following tests. The initial standing time was 5 minutes, the
meet the demand of clean oil defined by CIGRE working group time interval between two successive shots was fixed at 1
12.17 as the particle content with a diameter larger than 5 μm is minute. All experiments were performed at room temperature.
down to 300 per 100 ml in oil sample [10]. The particle number After one breakdown occurs, the oil sample and electrodes
in oil samples was decided by HIAC 8011 liquid particle were changed, totally 10 samples were carried out for each type
counting system, which can detect particles with diameter from of oils.
1 to 200 μm.
SNFs were prepared by dispersing TiO2 nanoparticles 2.4 PARTIAL DISCHARGE INCEPTION VOLTAGE
(average particle size diameter<20 nm) into transformer oil (PDIV) TESTS
with the volume percentage concentration of 0.075%. All the Partial discharge under ac voltages indicates localised
oil samples were degassed at less than 1 kPa for 2 days and the insulation defects, manufacturers endeavor to design
moisture content of each sample was measured using a /manufacture discharge free transformers. According to IEC
Metrohm 648 Coulometer according Karl Fischer titration 1294:1993, 50 mm gap of needle (25 mm long) with a 3 μm
method before test. radius of curvature to sphere electrodes was chosen in the
following tests. Voltage rising at a rate of 1 kV/s was used to
2.2 Measurement of Electric Parameter stress the samples. Considering the request of industry design,
Relative permittivity and dc resistivity of transformer oil and 10 breakdowns in IEC 1294 are not enough for such a purpose.
SNFs were measured according to IEC 60247-2004, using A 2 sets of 10 breakdowns were carried out for each oil sample.
Jiantong Automatic Oil Tester JDC-1. Totally, 20 breakdowns were obtained for each sample. The
breakdown voltage at 63.2% and 5% breakdown probability
2.3 Breakdown Voltage Tests were calculated by using a Weibull distribution.
A Portable Jiantong Oil Tester 6801 was used to measure ac After getting the PDIV, further investigations of PD
breakdown voltages as per ASTM D1816 using brass performance in nanofluid and mineral oil were carried out at 1
spherically-capped electrodes set at 2 mm gap. The rate of rise time and 2 times PDIV lasting 10 minutes with LDS-6 PD
of the voltage was 2 kV/s. The initial standing time of 5 minutes detector, respectively.
was fixed before the application of voltage. The time interval
with stirring action after each breakdown was set at 1 minute. 2.5 THERMALLY STIMULATED CURRENT (TSC)
All experiments were performed at room temperature. TESTS
Furthermore, considering the request of industry design, Trapping parameters of electrons in dielectric is determined
breakdown voltage at quite low breakdown probability is more by using the thermally stimulated current technique [11]. TSC
meaningful than the mean breakdown voltage. 6 breakdowns in is the current released from a pre-electrostressed dielectric as a
ASTM D1816 are not enough for such a purpose. 5 sets of 6 result of heating. The curve plotting the dependence of TSC on
breakdowns were carried out for all oil samples. Totally, 30 temperature can be used to calculate energy level of trap and
breakdowns were obtained for each sample. According to the trap density. The experimental was carried out using
large number of experimental data, the breakdown voltage at Novocontrol Liquid TSC equipment. The same experimental
63.2% and 5% breakdown probability were calculated by using setup as we used before [12] is shown in Figure 1. At
a Weibull distribution. atmosphere pressure, switch S3 was closed and a negative dc
A 10 stage impulse generator was used to provide 1.2 /50 μs field of 2 kV/mm was applied to the tested sample, the same
standard lightning impulse. According to ASTM D3300, 25 dielectric we used in experiments 1, for 20 min at 323 K. Then,
mm gap of needle with a 50 μm radius of curvature to sphere after the temperature was lowered down to 260 K, switch S3
772 Y. Du et al.: Effect of Semiconductive Nanoparticles on Insulating Performances of Transformer Oil

was opened and switch S1 was closed to discharge the Secondly, dc high voltage was powered off. Thirdly, the
equivalent capacitor of electrodes for one minute. Finally, change of space charge distribution is measured at time
switch S1 was opened and switch S2 was closed to begin the intervals of 8 s.
measurement of TSC by raising temperature at a rate of 2
K/min. A Keithley 6514 provides an accuracy of 10-16 A on the
3 RESULTS AND DISCUSSIONS
current measurement, and the accuracy of temperature
measurement is about ±0.2%. In order to ensure the accuracy of 3.1 RELATIVE PERMITTIVITY AND RESISTIVITY
test, TSC curves shown in Figure 5 were the average of three
measurements. Electric parameter values for both samples are given in Table
1. Though the resistivity values of SNFs in Table I show a
downshift with respect to the typical values of the transformer
oil, they still meet the quality of transformer oil in service.
However, the relative permittivity value of SNFs is much
higher than mineral oil which is beneficial in causing a more
uniform electric field in oil-paper insulation.
Table 1. Comparison of Relative Permittivity and Resistivity.
DC Resistivity
Samples Relative Permittivity
（Ω·m）
12
Mineral oil 2.26 1.82×10
10
Nanofluid 3.92 8.30×10

Figure 1. Experimental setup for measurement of thermally stimulated

currents. 3.2 AC BREAKDOWN VOLTAGE
The water content of all samples measured before tests
2.6 Pulse Electroacoustic (PEA) Tests was in the range of 9 ppm to 10 ppm. The results of increased
The charge distribution and decay in oil samples were measured transformer oil breakdown voltage with the addition of
by PEA. The setup of PEA experiment is shown in Figure 2. semiconducting nanoparticles for transformer oil and the
associated nanofluids are summarized in Table 2. It is observed
that the ac breakdown voltage achieves an increase of 20% by
the presence of nanoparticles. The coefficient of variation for
oil samples is relatively low indicating the breakdown for all
samples is more credible.

Table 2. Comparison of ac breakdown voltages between nanofluid and mineral

oil.
Samples Mean Breakdown S.D. Coefficient
voltage（KV） （KV） of variation

Mineral oil 67.9 5.4 8

Nanofluid 80.9 5.3 6.5

Figure 2. Experimental setup for pulse electroacoustic measurement.

Weibull distribution is the popular statistical methods used
to model the probability distribution of breakdown voltage
The lower electrode is made of 10 mm thick aluminum as the [13]. It is appropriate especially for calculating the withstand
anode. The upper electrode is made of 10 mm thick copper as voltages at low probabilities [14]. Therefore, we use Weibull
the cathode, attached a semiconducting polymer to achieve a distribution to calculate the withstand voltages and obtain the
better acoustic match. An acoustic transducer is attached on the shape parameter and scale parameter. The cumulated
lower electrode. The oil sample with 120 μm thickness is distribution function of Weibull distribution is given as
located between the two electrodes. A dc high voltage source is follow:
used to cause space charge in the sample. An electrical pulse x
-( )α
β
with 5 ns width and 600 V is applied to the sample for F(x) = 1 - e (1)
producing acoustic pressure waves by interaction between the
pulsed electric field and the charge layer. Detection of acoustic where, α is shape parameter, βis scale parameter and x is the
pressure waves can determine charge distribution across the breakdown voltage.
sample. During the experiment, firstly a negative dc high The modeling results of total 30 breakdowns for mineral oil
voltage was applied to the sample for 20 min. The initial space and nanofluid are shown in Figure 3. X-axis and Y-axis stand
charge measurement was done at the end of voltage application. for breakdown voltage and calculated probability respectively.
IEEE Transactions on Dielectrics and Electrical Insulation Vol. 19, No. 3; June 2012 773

It can be observed that the negative dc breakdown voltage

achieves an increase of 28%. Meanwhile, the positive dc
breakdown voltage of SNFs is a little less than that of
transformer oil. It shows that SNFs exhibits better insulation
performance in the condition where the electron emission is
violent.

3.5 PARTIAL DISCHARGE INCEPTION VOLTAGE

PDIV tests were carried out, and then Weibull distribution
was employed to analyze the results. As the results shown in
Table 6, PDIV at 5% probability of nanofluid rises to 1.18
times than that of pure mineral oil. The chance of partial
Figure 3. Weibull fitting of ac breakdown results.
discharge inception will be decreased by accession of TiO2
The withstand voltage at 63.2% and 5% breakdown nanoparticles.
probability are given in Table 3. The withstand voltage of Table 6. Results of PDIV.
nanofluid at 5% breakdown probability rises to 1.24 times than Samples Mean Standard 63.2% 5%
that of mineral oil, indicating the modification of TiO2 PDIV(KV) Deviation probability probability
(KV) PDIV (kV) PDIV (kV)
nanoparticles improves pure mineral oil’s reliability.

Table 3. Calculation of withstand voltage. Mineral

30.6 2.7 31.7 25.0
Samples 63.2% probability breakdown 5% probability breakdown oil
voltage (kV) voltage (kV) 33.1 1.8 33.8 29.5
Nanofluid
Mineral oil 69.5 56.9
Nanofliud 82.6 70.4

3.3 Lightning Impulse Breakdown Voltage

The results of impulse breakdown voltage for transformer oil
and the associated nanofluid are shown in Table 4. It is observed
that the impulse breakdown voltage achieves an increase of 24%.
It is interesting to notice that the time to breakdown of nanofluid
is prolonged by as much as 53% compared to that of the pure oil,
indicating a slower streamer velocity in SNFs. The results are
significant because a slower streamer requires more time to
traverse the gap between electrodes to cause breakdown. This
allows more time for the applied impulse voltage to be
extinguished. These results demonstrate that the presence of the
semiconducting nanoparticles in the oil effectively inhibits the
processes of electrical breakdown. (a) 1 time PDIV

Table 4. Lightning breakdown voltage of pure oil and nanofluid.

Samples Mean Standard Time to Standard
breakdown Deviation
Breakdown Deviation
(μs) (μs)
Voltage (kV) (kV)
Mineral oil 77.6 5.9 15.2 1.16
Nanofliud 95.9 8.6 23.3 0.34

3.4 DC BREAKDOWN VOLTAGE

The results of dc breakdown voltage for transformer oil and
the associated nanofluid are shown in Table 5.
Table 5. DC breakdown voltage of pure oil and nanofluid.

DC(+) Standard DC(-) Standard (b) 2 times PDIV

Samples Breakdown Deviation Breakdown Deviation Figure 4. Partial discharge phenomena under different PDIV value.
Voltage(KV) (KV) Voltage(KV) (KV)
Mineral
49.1 2.6 66.3 4.0
To investigate the resistance of insulating oil against PD, PD in
oil oil and nanofluid were carried out at 1 and 2 times PDIV for 10
Nanofluid 45.1 1.8 84.6 6.4
minutes. Figure 4 shows the partial discharge phenomena in all
774 Y. Du et al.: Effect of Semiconductive Nanoparticles on Insulating Performances of Transformer Oil

samples and the statistical results are summarized in Table 7. It obtained by spectroscopy [16]. It is interesting to note that on
appears clearly that there is a significant increase of PD magnitude the TSC curves the peak value of TSC curve for SNF is 2.5
and impulse number in transformer oil than those of SNFs. With times of that for pure oil. Since the peak value of TSC curve is
the increase of applied voltage, the discharge phenomenon gets related to maximum trap density in a dielectric liquid, SNF is
more violent and the modification effect of nanoparticles on the assumed to have a higher trap density than the pure oil. The
PD character of mineral oil becomes more distinct. integration of TSC over temperature gives the total number of
Table 7. Statistical Results of PD Performance trap charge released from the trap center. It is found that the
1 time PDIV 2 times PDIV
total number of charge trapped in nanofluid is 6.057 nC, 1.85
Mineral Mineral
times of that trapped in the pure oil. These results demonstrate
nanofluid nanofluid that the presence of TiO2 nanoparticles in the oil dramatically
oil oil
increases the electron trap density at shallow trap levels.
Pulse number 9 6 245 168
To further investigate the charge carriers transport
Mean discharge characteristics in the dielectric liquid, the charge distribution
435 245 6062 5180
magnitude (pC) and decay in oil samples were measured by PEA. The
experimental results are shown in Figure 6. The positions of
4 POSSIBLE MECHANISMS electrodes were confirmed and marked with black lines. The
Our experimental results demonstrate that TiO2 anode peak is sharp and evident. In contrast, the cathode peak is
nanoparticles with long relaxation time have significantly wide and flat, because of attenuation and scattering of acoustic
improved the breakdown strength of transformer oil. However, wave caused by the oil sample. It is noted that the space charge
the present model of electron scavengers judged by the charge existing in SNF (1.1 C/m3) is almost two times of that in pure
relaxation time constant τ for the transformer oil/nanoparticle oil (0.6 C/m3) when the dc is applied. It is also interesting to
system [7], fails to explain the modification mechanism of TiO2 note that the maximum charge density decay velocity for SNF
nanoparticle on the breakdown strength of transformer oil. is 4 times of that for pure oil in the first 8 seconds after the
voltage is removed. This rapid charge dissipation was also
observed in a dielectric nanocomposites modified by TiO2
(2) nanoparticles [17].
where the spherical nanoparticle permittivity ε2 and
conductivity σ2, and is surrounded by transformer oil with
conductivity σ1 and permittivity ε1.
It is well known that electrical breakdown in transformer oil
is closely related to the electron transport process, which
includes electron hopping transport in the traps and transport in
a delocalized state [15]. Therefore, we measured the trap
characteristics of SNF and pure transformer oil by the TSC
method in order to understand the effect of TiO2 nanoparticles
on trap level and trap density of the oil samples.

Figure 5. The results of thermally stimulated current measurement.

As shown in Figure 5, trap energy levels of the pure

transformer oil and SNF are both at shallow trap levels. The
trap center for the pure oil and SNF are calculated to be 0.416
eV and 0.418 eV, respectively. This observation is in
agreement with the shallow trap states of TiO2 nanoparticle Figure 6. Accumulation and decay of space charge distribution characteristics.
IEEE Transactions on Dielectrics and Electrical Insulation Vol. 19, No. 3; June 2012 775

The rapid charge dissipation in SNF could be attributed to nanofluids”, IEEE Conf. Electr. Insul. Dielectr. Phenomena (CEIDP), pp.
61-366, 2008.
high shallow trap density created by TiO2 nanoparticles [18],
[7] J. G. Hwang, M. Zahn, F. O'Sullivan, L.A. A. Pettersson, O. Hjortstam
and associated with charge hopping transport process in the and R. Liu, “Effects of nanoparticle charging on streamer development in
shallow traps. Based on our test results of TSC and PEA, it is transformer oil-based nanofluids”, J. Appl. Phys, Vol.107, 014310, 2010
proposed that the fast electrons created at high field in the oil [8] P. Kopcansky, L.Tomco, K.Marton, M.Koneracka, I.Potocova and M.
can be captured by the shallow traps and released rapidly from Timko, “The dc dielectric breakdown strength of magnetic fluids based on
transformer oil”, J. Magnetism and Magnetic Materials, Vol. 289, pp.
these shallow traps in the oil. The fast electrons can be 415–418, 2005.
converted to slower electrons by repeated trapping and [9] M. Chiesa and S. K. Das, “Experimental investigation of the dielectric and
de-trapping in the oil [19], when moving from high field to low cooling performance of colloidal suspensions in insulating media”,
field locations. Electron trapping and de-trapping process in the Colloids and Surfaces A: Physicochem. Eng. Aspects, Vol. 335, pp.
88-97, 2009.
shallow traps could be one of the main charge transport [10] CIGRE, "Effect of particles on transformer dielectric strength", Working
processes in dielectric liquids. High shallow trap density Group 17 of Study Committee 12, June 2000.
created by TiO2 nanoparticles is anticipated to greatly [11] R. Chen, Analysis of Thermally Stimulated Processes, Pergamon, Oxford,
contribute into the reduced speed of the streamer propagation UK, 1981.
[12] M. Li, C. Li, M. Zhan, and J. Xu, “Effect of surface charge trapping on
and the improved dielectric strength of SNF. dielectric barrier discharge”, Appl. Phys. Letts., Vol. 92, 031503, 2008.
[13] Q. Liu, Z.D. Wang, “AC and lightning breakdown strength of mineral oil
CONCLUSIONS Nytro Gemini X and 10GBN”, Int. Electr. Insul. Conf. (INSUCON),
Birmingham, UK, pp. 26-28, 2009.
(1) Semiconductive nanoparticles can increase the ac, dc and [14] D. Martin and Z.D. Wang, “Statistical analysis of the ac breakdown
lightning impulse breakdown voltages of SNFs up to more voltages of ester based transformer oils”, IEEE Trans. Dielectr. Electr.
Insul., Vol. 15, pp. 1044-1050, 2008.
than 1.2 times compared with pure oils. Especially, the [15] R. Bartinikas, Electrical Insulating Liquids, ASTM Press, Philadelphia,
time to impulse breakdown of SNFs is prolonged by as pp. 262-307, 1994.
much as 53%. [16] H. Takahashi, R. Watanabe, Y. Miyauchi and G. Mizutani, “Discovery of
(2) Both shallow trap density and charge dissipation velocity deep and shallow trap states from step structures of rutile TiO2 vicinal
surfaces by second harmonic and sum frequency generation
in SNFs are greatly increased compared to those in pure spectroscopy”, J. Chem. Phys., Vol. 134, 154704, 2011.
transformer oils by 2.5 times and 4 times, respectively. [17] R. C. Smith, C. Liang, M. Landry, J. K. Nelson and L. S. Schadler, “The
(3) It is proposed that electron trapping and de-trapping mechanisms leading to the useful electrical properties of polymer
nanodielectrics”, IEEE Trans. Dielectr. Electr. Insul., Vol. 15, pp.
process in the shallow traps could be one of the main 187-196. 2008.
charge transport processes in transformer oil. The [18] J.K. Nelson and J.C. Fothergill, "Internal charge behavior of
increased shallow trap density in the SNFs could convert nanocomposites", Nanotechnology, Vol. 15, pp. 586-595, 2004.
fast electrons to slow electrons by trapping and de-trapping, [19] K. Wu and L. A. Dissado, “Model for electroluminescence in polymers
during the early stage of electrical tree initiation”, IEEE Int’l. Conf. Solid
offering the significant improvement of breakdown
Dielectr., Vol. 2, pp.505-508, 2004.
performance of SNFs.
Yue Fan Du was born in Tianjin China., in 1986. He is
an electric power engineering graduate student at North
ACKNOWLEDGMENT China Electric Power University. He received the B.S.
degree in electrical engineering from North China
We thank the National Natural Science Foundation of China Electric Power University in 2008. His research interests
for supporting this research under Contract No. 51077050, the include nanocomposite high voltage insulation material
Chinese Universities Scientific Fund, the Research Project of and condition monitoring of power apparatus.
State Grid Corporation of China and the National Engineering
Laboratory for Ultra High Voltage Engineering Technology
(Kunming, Guangzhou). Yu Zhen Lv received the Ph.D. degree from the Beijing
University of Aeronautics and Astronautics in Materials
REFERENCES Science and Engineering. She is currently an Associate
Professor of Materials Science and Engineering at North
[1] IEEE, Guide for loading Mineral-oil-immersed Transformers, Annex I: China Electric Power University. Her research interests
Transformer Insulation Life, IEEE Standard C57.91, 1995. are in synthesis, characteristic and application of
[2] D. Peterchuck and A. Pahwa, “Sensitivity of transformer’s hottest-spot nanodielectrics.
and equivalent aging to selected parameters”, IEEE Trans. Power
Delivery, Vol.17, pp.996-1001, 2002.
[3] EPRI Portfolio 2012- Transmission reliability and performance: 37.002, Cheng Rong Li (SM’03) was born in Xian, China on 1
transformer life extension. http://www.epri.com/portfolio/ March 1957. He received the B.S. and M.S. degrees in
[4] V. Segal, A. Hjortsberg, A. Rabinovich, D. Nattrass and K. Raj, “AC (60 electrical engineering from North China Electric Power
Hz) and impulse breakdown strength of a colloidal fluid based on University in 1982 and 1984, respectively. He received
transformer oil and magnetite nanoparticles”, IEEE Int’l. Sympos. Electr. the Ph.D. degree in electrical engineering from Tsing Hua
Insul. Arlington, VA, USA, pp. 619-622, 1998. University in 1989. He joined the University of South
[5] V. Segal, A. Rabinovich, D. Nattrass, K. Raj, and A. Nunes, Carolina in USA in 1992 as a postdoctoral research fellow.
“Experimental study of magnetic colloidal fluids behavior in power He joined North China Electric Power University
transformers”, J. Magnetism and Magnetic Materials, Vol. 215, pp. (NCEPU) in 1995. Now he is a Professor in the
513-515, 2000. Department of Electrical Engineering of NCEPU. He is a fellow of IEE. His
[6] J. G. Hwang, O'Sullivan, M. Zahn, O. Hjortstam, L.A.A. Pettersson and research interests include gas discharges, electrical insulation and materials and
R. Liu, “Modeling of streamer propagation in transformer oil-based condition monitoring of power apparatus.
776 Y. Du et al.: Effect of Semiconductive Nanoparticles on Insulating Performances of Transformer Oil
Mu Tian Chen was born in Hunan China, in 1987. He is an Xiao Xin Li was born in Baoding China in 1986. She is a
electric power engineering graduate student at North China power engineering graduate student at North China
Electric Power University. His research interests include Electric Power University. She received the BS. degree
high voltage insulation material and condition monitoring in physical science and information engineering from
of power apparatus Hebei Normal University in 2009. Her research interests
are synthesis of nanomaterials, and application in
insulation.

Yu Xiang Zhong was born in Sichuan China, in 1988.He

is a control and computer engineering graduate student at
North China Electric Power University. His research
interests include high voltage insulation material.
You Zhou was born in 1987, in Hunan, China and
received the B.S. degree in electrical engineering from
South China University in 2009. He is an electric power
engineering graduate student at North China Electric
Power University. His research interests include
Jian Quan Zhou was born in Shandong China in 1987. nanocomposite high voltage insulation material and
He is an electric power engineering graduate student at space charge theory.
North China Electric Power University. His research
interests include nanocomposite high voltage insulation
material and power system protection.