IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 21, NO.

3, JULY 2006

1291

Earthquake Simulator Testing of

Base-Isolated Power Transformers
Nobuo Murota, Member, IEEE, Maria Q. Feng, and Gee Yu Liu

AbstractThis paper presents a comprehensive study involving

triaxial earthquake simulator testing on seismic isolation of electric
power transformers. In this study, two isolation systems were developed: one using sliding bearings combined with rubber bearings
and the other segmented high-damping rubber bearings. Triaxial
earthquake simulator testing was performed using a large-scale
transformer model equipped with a real bushing. The effectiveness
of the base isolation system in reducing the response acceleration
of power transformer systems was demonstrated. Important observations were made on the seismic responses of the transformer and
bushing. In particular, the vertical component of the ground motion induced a high-frequency response of the bushing when the
transformer was isolated with the sliding isolation system. This is
because the vertical motion changes the friction forces in the sliding
bearings that excite high modes in the transformer-bushing system.
Furthermore, the effect of the interaction with the bushing connecting cables on the response of the bushing in the base-isolated
system was experimentally evaluated. In conclusion, the base isolation technology, when properly designed, is a highly effective measure for seismic protection of power transformers.
Index TermsBase isolation, bushing, earthquake simulator,
electric power transformer, seismic, substations.

I. INTRODUCTION

ECENT destructive earthquakes including the 1994

Northridge Earthquake in the United States, the 1995
Kobe (Hyogo-ken Nanbu) Earthquake in Japan, the 1999 Izmit
Earthquake in Turkey, and the 1999 Chi-Chi Earthquake in
Taiwan, all caused significant damage to the electrical power
networks, important lifelines that deliver electric power to
residents and businesses in urban and suburban areas. While the
duration of system disruption was relatively short to moderate,
one day for Northridge (1995) [1]; three days for Kobe (1995)
[2]; and two weeks for Chi-Chi, the direct losses were estimated
to be in the range of hundreds of millions of dollars for each
event.
Substations are one of the key facilities in an electrical
power network. Older substation equipment, particularly power
transformers, are expensive and were designed to much lower
seismic standards or none at all and are vulnerable to seismic
damage. Many transformers suffered severe damage during
Manuscript received February 3, 2005; revised September 24, 2005. Paper
no. TPWRD-00055-2005.
N. Murota is with the Department of Seismic Isolation Engineering,
Bridgestone Corporation, Yokohama 244-8510, Japan (e-mail: murota-n@
bridgestone.co.jp).
M. Q. Feng is with the Department of Civil and Environmental Engineering,
University of California at Irvine, Irvine, CA 92697-2175 USA (e-mail:
mfeng@uci.edu).
G. Y. Liu is with the National Center for Research on Earthquake Engineering,
Taipei 106, Taiwan, R.O.C. (e-mail: karl@ncree.gov.tw).
Digital Object Identifier 10.1109/TPWRD.2006.874586

the past earthquakes including anchorage failure ripping the

transformer case and causing oil leakage, as well as the fracture
of porcelain bushings and damage in the connection of the
bushing and transformer. Damage also occurred inside the
transformer bodies that are usually difficult to repair and a few
months are needed to replace the transformer [3].
The base isolation technology has gained popularity in this
decade as one of the rehabilitation measures for seismic protection of building and bridge structures. Fujita et al. (1984) [4]
performed an experimental and analytical study on base isolation of equipment, such as a power transformers, focusing on
the development of isolation devices. While this study demonstrated the effectiveness of the base isolation, it did not consider
the interaction of porcelain bushings with the transformer. In
fact, their test frame did not contain any bushings. G. Bonacina
et al. (1995) [5] performed several feasibility studies of base
isolation for substation facilities by numerical simulation, introducing design examples of base isolation systems for 170-kV
gas-insulated substations. Ersoy and Saadeghvaziri (2004) [6]
carried out an analytical study on the seismic response of transformer-bushing systems with the finite-element model focusing
on the interaction between the transformer and bushing. They
concluded that the transformer tank flexibility affects bushing
dynamic characteristics, and this effect is mainly due to the flexibility of the top plate.
The objective of this research is to investigate the effectiveness and applicability of the base isolation technology for
seismic protection of power transformers through experimental
study using a large-scale transformer model with real bushings
by comparison with the fixed-base system. Triaxial earthquake
simulator testing with a large-scale model was used for the first
time in this field.
The entire testing program was part of a joint research
project with the National Center of Research on Earthquake
Engineering (NCREE) in Taipei, Taiwan, R.O.C., and the earthquake simulator test was conducted at NCREE. A large-scale
transformer-model, bushing, and isolation system were designed by authors and manufactured by Bridgestone Corporation Japan for the project. The testing program was classified
into three phases. In Phase-1 testing, a frame-structure with lead
2 1.8-m height was
counter weights for 235.5 kN, of 2
built to model the transformer body. Actual porcelain bushing
of 161 kV was mounted on the frame-structure. An isolation
system consisting of sliding bearings and low-damping rubber
bearings was designed and subjected to the testing [7]. In
Phase-2 testing, a segmented high-damping rubber bearing
system (SHRB) was applied as the isolation system. The same
frame structure with a 145-kN counter weight was used for
this testing. A flexible rubber ring was developed and applied

0885-8977/$20.00 2006 IEEE

1292

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 21, NO. 3, JULY 2006

TABLE I
NATURAL FREQUENCY OF TRANSFORMER AND BUSHING

TABLE II
DESIGN PARAMETERS OF ISOLATION SYSTEM

Fig. 1. Transformer model and bushing.

in the test to investigate the difference in the bushing response

with a low frequency type. In Phase-3 testing, a preliminary
study of the interaction problem in a base-isolated transformer
was evaluated. The dynamic interaction of the cable-connected
element in the substation may cause damage in the equipment
during an earthquake [8]. The transformer in Phase-2 was cable
connected to a pole, fixed on the platform of the simulator, and
then subjected to uniaxial shaking. An aluminum strand cable,
which had been previously used in a substation, was connected
from the top of the bushing to the pole top.
II. EARTHQUAKE SIMULATOR TESTING WITH
SLIDING BEARING SYSTEM/PHASE-1
A. Earthquake Simulator
The size of the earthquake simulator is 5 5 m in plane and
the maximum payload is 500 kN. The simulator has six degrees
mm
of freedom and the maximum stroke and velocity are
and 1000 mm/s, respectively.
B. Transformer Model
Because of the limitation of the payload capacity of the earthquake simulator, the transformer model was scaled down with
a weight of 235.5 kN. The model is a four-layer steel frame
structure, where lead blocks were loaded to provide additional
weights. The main body of the transformer is assumed as a
single mass in this model. At the top of the frame, a bushing
is connected with bolts to the mounting plate. The top plate of
the transformer model, where the bushing is mounted, is also
rigid compared with real transformers. Fig. 1 shows the entire
view of transformer model and bushing. The 161-kV bushing
was used in the testing. Their natural frequencies were evaluated by random wave excitation, which are shown in Table I.
The equivalent damping ratio of each components was around
2%.
C. Isolation System
Four sliding bearings were installed at the corners of the
transformer model and two rubber bearings at the middle of
the two corners. Sliding bearings carry the entire weight of the
transformer model, including the bushing, and rubber bearings

work as horizontal restoring force elements without sustaining

any vertical load. In order to reduce the stiffness and maintain
a large deflection capability, each rubber bearing unit consists
of two rubber bearings of 304 mm in diameter and 96-mm total
rubber height, which is double decked and fixed with bolts
through flange plates of each upper and lower bearing. A sliding
bearing consists of a laminated rubber pad of 120-mm diameter
and a PTFE disc of 55-mm diameter fixed together with keys.
The total parameter values of isolation system are summarized in Table II.
D. Instrumentation
Accelerometers and displacement transducers for , , and
directions were installed at the simulator platform, at the bottom
and top of the transformer model, and at four points on the
bushing at the following locations (measured from the middle
flange of the bushing).
cm, second node
cm, third node
first node
cm, fourth node 197.5 cm.
Three-component load cells for the measurement of the reaction
forces of the sliding and rubber bearings were installed above
each bearing. The detailed information of the instruments and
their locations is shown in Fig. 2.
E. Testing Program
, and
The earthquake simulator was excited in the ,
directions. The direction was set as the main direction for excitation, and the component with the largest phase
grid array (PGA) of each motion was set as the x-direction excitation. The other components in the horizontal direction were
set as the -direction motion and the updown component as the
-direction motion. The earthquake records and its PGA were as
follows.
Ground motion: 1940 El Centro, 1994 Northridge
(Sylmar), 1995 Kobe (Takatori);

MUROTA et al.: EARTHQUAKE SIMULATOR TESTING OF POWER TRANSFORMERS

Fig. 2. Test setup and instrumentations.

Fig. 4. Maximum response acceleration: Kobe/x375.

Fig. 3. Maximum response acceleration: Northridge/x375.

PGA: uniaxial

1293

Fig. 5. Total force-displacement curve of isolation system in Kobe/x375.

biaxial
triaxial

F. Test Results
Figs. 3 and 4 plot the distribution of the response acceleration
at different measurement points including platform, bottom, top
of the transformer, and top of the bushing, under Northridge and
Kobe.
The effectiveness of base isolation is evident in the plots.
In the base-isolated system, the amplification of the response
acceleration is very small from the bottom to the top of the
transformer. The total force-displacement loops of the isolation
system, including the rubber and sliding bearings, are plotted in
Fig. 5 by superposition of the force-displacement loops of all
the bearings.
Under the triaxial shaking, the responses of the power transformer showed a significant difference from those under uni
and biaxial shaking. The response acceleration at the top of

the bushing was amplified and, in some cases, the response exceeded that of the fixed-base system. Fig. 6 compares the peak
response acceleration along the height of the transformer under
uni (x375), bi (xy-375), and triaxial shaking (xyz-375) in the
cases of Kobe (Takatori). This phenomenon was a very significant finding through Phase-1 testing. Fig. 7 shows the force-displacement loop of the sliding bearing under triaxial shaking.
The loops under triaxial shaking were obviously affected by the
change of the vertical load due to the vertical excitation. From
the above discussion, it was deduced that the high-frequency
factor on the friction force of the sliding bearings, which was
caused by the vertical excitation, affected the bushing response.
III. EARTHQUAKE SIMULATOR TESTING WITH HIGH DAMPING
RUBBER BEARING SYSTEM/PHASE-2
A. Segmented High-Damping Rubber Bearing
Fig. 8 shows the segmented high-damping rubber bearing
(SHRB) used in this test. The isolation system consisted of three
stacks of four bearings with a diameter of 72 mm, which are
called element bearings. The thick plates between each bearing

1294

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 21, NO. 3, JULY 2006

Fig. 6. Comparison of maximum response acceleration in uni, bi, and triaxial

shaking Kobe/x, xy, xyz375.

Fig. 8.

Segmented high-damping rubber bearing (SHRB).

TABLE III
DESIGNED PROPERTIES OF SHRB

Fig. 7. Force-displacement curve of sliding bearing under triaxial shaking

Northridge/xyz375.

layer were designed to work as stabilizers during large displacement [9]. The designed ultimate shear strain of each element
bearing, which is decided by buckling, is around 250%. The
total maximum displacement of SHRB was around 200 mm.
The nominal compressive stress was 4.0 MPa. The design shear
modulus and equivalent damping ratio at 100% shear strain was
0.61 MPa and 16%, respectively. The diameter of the element
bearing was 72 mm and the thickness of the unit rubber layer
was 0.9 mm. The number of layers was 31 and the total rubber
height was 27.9 mm. The design natural period of the transformer model sustained by the four SHRB was computed as
1.32 s. The design properties are summarized in Table III.
B. Flexible Rubber Ring
According to the field testing of Villaverde [10], the natural
frequency of 500-kV bushings in the substation, which includes
the influence of the stiffness of turrets and top plates of transformers, was from 3 to 4 Hz. Also, Ersoy et al. [6] indicated by

analytical study that the bushing dynamic characteristics are affected by the flexibility of the top plate of the transformer. On
the other hand, the 161-kV bushing in the Phase-1 test has a
relatively higher frequency, 12.5 Hz, compared with the large
size bushing. Therefore, in order to evaluate the response of the
bushing with low natural frequency, such as 3.0 Hz, a flexible
rubber ring (shown in Figs. 9 and 10) was specially designed and
manufactured, and was mounted between the top of the turret
and the flange of the bushing. The rubber ring was designed to
contribute to the rocking motion of the bushing and shift the fundamental period of the bushing with its low tilting stiffness. The
of the rubber ring is calculated
approximate tilting stiffness
by the following equation [4]:
(1)
(2)

MUROTA et al.: EARTHQUAKE SIMULATOR TESTING OF POWER TRANSFORMERS

1295

TABLE IV
DESIGN PARAMETERS OF FLEXIBLE RUBBER RING

Fig. 9.

Fig. 10.

Flexible rubber ring.

Cross-section view of a flexible rubber ring.

where
Youngs modulus of rubber for bending;
Youngs modulus of rubber (
: shear modulus)
MPa;
bulk modulus of rubber 1200 MPa for this compound;
first shape factor of rubber ring;
correction factor 0.85 for this compound;
rubber-layer thickness;
section modulus of the ring.
A low damping rubber compound was used, with a shear
modulus of 0.4 MPa. The design characteristics of the rubber
ring are shown in Table IV.
C. Testing Program
The differences between the Phase-2 and the Phase-1 testing
programs were the isolator system, the input ground motion, the
total weight of the model, the application of the flexible rubber
ring, and the application of a scale factor. The scale factor applied was 0.6 for length and displacement. In order to maintain
the scale factor for acceleration stress as 1.0, the time step was
reduced for 0.6
.
The parameters of the earthquake simulator tests were 1) base
isolated or fixed base; 2) with rubber ring (low-frequency mode)
or without rubber ring (high-frequency mode); 3) earthquake
record; 4) intensity of earthquake record; and 5) direction
of shaking. The earthquake records used in Phase-2 were
1995 Kobe (Takatori), 1999 Chi-Chi (TCU-129), and Art-693.
Art-693 was the artificially generated wave based on the required response spectrum (RRS) IEEE-693 [11]. The phase

Fig. 11.

Response spectra of artificial-wave (N-S component).

angles of the composed waves were randomly chosen and superimposed. The comparison of design response spectra and the
response spectra of the generated artificial wave in 2% damping
is shown in Fig. 11. The typical dynamic characteristic of
Art-693 is the high intensity in the low-frequency range, similar
to 1995 Kobe (Takatori), close to the fundamental frequency
of the base-isolated transformer. The shaking direction was
uniaxial in the direction; biaxial in the and directions;
triaxial in the , , and directions; the same as Phase-1.
biaxial shaking in the x and z directions was added in Phase-2.
The combination of the record, intensity, and shaking direction
in each test case is the same as that in Phase-1.
The dynamic identification of the transformer model and
bushing was performed by a random vibration test using the
same procedure as in Phase-1. As predicted, the results were
almost the same as Phase-1. The first and second mode of the
transformer model without the bushing was around 16.4 Hz,
direction. The first and second
and 29.6 Hz in the , , and
mode of the 161-kV bushing without the rubber ring was 12.7
and 16.3 Hz, whereas with a rubber ring, it was as follows:

1296

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 21, NO. 3, JULY 2006

Fig. 12. Maximum response acceleration at each measurement point:

Art/x375.

first mode 3.9 Hz; second mode

26.1 Hz.

Fig. 13. Comparison of maximum response acceleration in uni, bi, and triaxial
shaking : B r /Art-693.

12.6 Hz; and third mode

D. Test Results
The notation of each system hereafter is as follows:
base-isolated with
; base isolated without;
; fixed-base with ring;
; fixed base;
ring
.
without ring
Fig. 12 shows the distribution of the response acceleration
and response relative displacement along the height of the
system under Art-693/x375. The result shows the interaction
effect between the bushing and the transformer model. In the
system, the fundamental frequency in the transformer (16
Hz) was close enough to that of the bushing (12 Hz) to cause
system, they were almost
amplification, while in the
decoupled since the fundamental frequency of the bushing was
reduced to 3.9 Hz by the rubber ring. According to a survey by
this author, the fundamental frequency of a transformer body
generally varies from 15 to more than 30 Hz. The results in this
study indicate that the relationship of the bushing frequency to
the transformer frequency has a significant influence in the amplification of the response in the bushing and will be one of the
major reasons for severe damage in bushings. The results in the
and
systems show good reduction in acceleration.
There was no obvious difference in the accelerations between
and
systems, regardless of the difference in
the
the fundamental frequency of the bushing. It is because that the
response of the system in base isolation frequency is dominant
and the bushing response is not sensitive to its natural frequency.
This result indicates the efficacy of the designed base isolation
system for bushing with a low natural frequency. Fig. 13 shows
the comparison of response acceleration in uni, bi, and triaxial
/Art-693. There was no significant difference
shaking of
in the response. The maximum displacement/shear-strain experienced during the entire test program was 199 mm/178% in
/Kobe/x375, as shown in Fig. 14. The predicted ultimate
shear strain of the SHRB was around 200%, at which the ratio

Fig. 14.

Shear stress-strain curve of isolation system in B

0 r/Kobe/x375.

of the diameter to the displacement was 0.8. In real scale, this

199 mm translates to 199/0.6 = 331 mm. Considering many
cable connections in actual transformer systems, this response
displacement was controlled in a reasonable range. A major
difference between the base isolation for transformer systems
versus conventional isolation, such as building isolation, is the
higher stiffness for limiting the displacement in a transformer.
As a result, the isolation period will be less than 1.5 s, whereas
the period of a conventional system is generally over 2.5 s.
The initial purpose of applying this rubber ring was to reduce
the natural frequency of the bushing and to evaluate the base-isolation effect in the flexible bushing system. However, the results
-system)
in the fixed-base system with the rubber ring (
indicated that this flexible-joint system itself had a significant
effect in reducing the response of acceleration in the bushing of
a fixed-base system. The rubber ring can be one of the effective measures in improving the seismic performance of power
transformers.

MUROTA et al.: EARTHQUAKE SIMULATOR TESTING OF POWER TRANSFORMERS

Fig. 15.

1297

Cable-connected base-isolated transformer model.

Fig. 17. Maximum response acceleration at each measurement point in the
fixed-base and the base-isolated system

was assumed as less than 200 mm according to the results in

the Phase-2 testing. Therefore, the cable had enough length to
accommodate the movement of the base-isolated transformer
model without hardening. An H-beam measuring 200 200
16 mm was used as the pole fixed to the platform of the simulator. The natural frequencies of the pole with a clamped-free
condition were calculated as 9.18 Hz in the first mode, and
57.5 Hz in the second mode. The load generated in the cable is
not measured.
B. Test Program

Fig. 16.

Geometry of test setup.

IV. RESPONSE OF CABLE-CONNECTED

TRANSFORMERS/PHASE-3
With additional testing, the effect of the cable connection
to the response of the base-isolated transformer model and the
bushing was preliminarily investigated in Phase-3 testing.
A. Test Setup
Figs. 15 and 16 show the experimental setup. The specification of the cable was 954MCM AAC (code word Magnolia
in the ASTM B231) and its material was pure aluminum of
E. C. Grade. The diameter was 28.55 mm and the tensile
strength was 7420 kg. The stranding was of Class AA and
37 4.079-mm diameter. The length of the cable was 2660 mm
and the horizontal distance from the pole to the bushing was
1880 mm. The horizontal displacement of the transformer to
where the cable would start to harden and pull the pole was
calculated from the geometry of Fig. 15 as 493.9 mm. The
maximum displacement of the base-isolated power transformer

The earthquake records and the PGA are summarized as

follows:
Ground motion: Art-693, 1995 Kobe (Takatori), 1999
Chi-Chi (TSU-129);
PGA: uniaxial
g
g
The 161-kV bushing was installed on to the transformer
model without the use of a rubber ring. The transformer with
and without an SHRB base isolation system was subjected to
the uniaxial shaking.
C. Test Results
The maximum response acceleration at each measurement
point in the fixed-base and base-isolated system is shown in
Fig. 17. The effect of base isolation is obviously seen. In Fig. 18,
the results of the response acceleration in the base-isolated
system, with and without the cable connection that was obtained
in Phase-2 testing, are compared. The response acceleration of
the bushing top with the cable connection was amplified 1.5 to
2.0 times the acceleration at the transformer top.
The response displacement at each measurement point in the
base-isolated system is shown in Fig. 19. As much as the response displacement of the transformer is large, the amplification of response acceleration of the bushing becomes high.
Fig. 20 shows the Fourier amplitude of response acceleration at
the bushing top and the pole top. The pole has its dominant frequency at around 14.6 Hz. The Fourier amplitude at the bushing

1298

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 21, NO. 3, JULY 2006

Fig. 20. Normalized Fourier amplitude at the bushing top and pole top in the
base-isolated system under Art/x375.
Fig. 18. Maximum response acceleration at each measurement point in the
base-isolated system

Fig. 19. Maximum response displacement at each measurement point in the

base-isolated system

top also has a second peak around 14 Hz corresponding to the

dominant frequency of the pole. From these facts, it is assumed
that the stiffness, even low, of connected cables in horizontal direction affected the dynamic response of the bushing top during
shaking.
V. CONCLUSION
The objective of this research was to experimentally study the
applicability of base isolation technology for seismic protection
of a transformer. The triaxial earthquake simulator testing was
carried out with a large-scale transformer model with real porcelain bushing. The transformer body was assumed to be a single
mass. So the transformer model was a rigid steel frame structure. The testing program consisted of two phases. In Phase-1,

the combination of sliding bearings and low-damping bearings,

was developed and applied as an isolation system. The results
in uni and biaxial shaking clearly proved that base isolation reduces the response acceleration of the transformer and bushing.
However, under triaxial shaking, the response acceleration at the
bushing was not reduced and, in some cases, it was amplified.
This was a very significant finding in Phase-1 testing. It was deduced that the vertical ground motion causes the high-frequency
load change on the friction force of the sliders, and it stimulated
the response of acceleration. This phenomenon was not indicated in any past research conducted in this field.
In Phase-2 testing, segmented high-damping rubber bearings
were developed and applied as an isolation system. The results
verified the effectiveness of base isolation, in the same way as
the slider system in Phase-1. Furthermore, there was no amplification of the bushing in the triaxial shaking, which was
observed in Phase-1. There was another significant development in the Phase-2 test. The rubber ring was specially designed
as a flexible joint for the bushing in order to evaluate the response of the bushing with low natural frequency. The natural
frequency of the bushing was shifted from 12.5 to 3.9 Hz with
the rubber ring. As a result, in the base-isolated system, there
was no obvious difference in the accelerations between with
and without the rubber ring systems, regardless of the difference in the fundamental frequency of the bushing. This fact indicates that base isolation is effective for the transformer-bushing
system regardless of the dynamic characteristics of the bushing.
In the fixed-base system, the response of the bushing was reasonably reduced.
In Phase-3, as additional testing, the base-isolated power
transformer with a cable connection was subjected to a uniaxial
shaking test to preliminarily investigate the interaction between
the bushing and the groundfixed pole. Although the results
show some effectiveness of the base isolation in comparison
with the fixed-base system, an interaction exists between the
bushing and the pole. Enough attention should be paid to
the relationship between the transformer and other connected
facilities in the design of base isolation systems. The test results
prompt further study on this problem.

MUROTA et al.: EARTHQUAKE SIMULATOR TESTING OF POWER TRANSFORMERS

ACKNOWLEDGMENT
Support for this research has been provided by the Multidisciplinary Center for Earthquake Engineering Research (MCEER).
The earthquake simulator testing was carried out at the National
Center for Research on Earthquake Engineering (NCREE) in
Taiwan.
REFERENCES
[1] Intermediate Term Plan for Seismically Hardening the Los Angels
Transmission-Level Power Facilities, 1995. Power Syst. Seismic Program Manage. Comm. City of Los Angeles Dept. Water and Power.
[2] M. Shinozuka, The Hanshin-Awaji Earthquake of January 17, 1995
Performance of Lifelines, National Center for Earthquake Engineering
Research, State Univ. New York, Buffalo, Tech. Rep. NCEER-95-0015,
1995.
[3] Base Isolation Support of Heavy Equipment with Laminated Rubber
Bearings1, 1984. in Japanese.
[4] Multidisciplinary Center for Earthquake Engineering Research
(MCEER). Research Progress and Accomplishments: 2000-2001.
[5] G. Bonacina, P. Bonetti, A. Martelli, F. Bettinali, and G. Serino, Seismic
base isolation of gas insulated electrical substations: Design, experimental and numerical activities, evaluation of the applicability, in Proc.
10th Eur. Conf. Earthquake Engineering, Duma (ed.), 1995.
[6] S. Ersoy and A. Saadeghvaziri, Seismic response of transformerbushing systems, IEEE Trans. Power Del., vol. 19, no. 1, pp. 131137,
Jan. 2004.
[7] N. Murota and M. Q. Feng, Hybrid base-isolation of bushing-transformer systems, Proc. ASCE, May 2123, 2001.
[8] A. D. Kiureghian, J. L. Sackman, and K.-J. Hong, Interaction in Interconnected Electrical Substation Equipment Subjected to Earthquake
Ground Motion, Pacific Eng. Res. Ctr. (PEER), Report PEER 1999/01,
Feb. 1999.
[9] N. Masaki, Study of Multistage Rubber Bearings for Seismic Isolation and Vibration Control System, D.Eng. dissertation, Univ. Tokyo,
Tokyo, Japan, 1999. (in Japanese).
[10] R. Villaverde, Ground Motion Amplification at Base of Bushings
Mounted on Electric Substation Transformer, Dept. Civil Eng., Univ.
California, Irvine, CA, Tech. Rep. Res. Supported by PEER/PG&E
Under Award PGE-09 566, 1999.

1299

[11] Recommended Practices for Seismic Design of Substations, 1998. IEEE

Std. 693-1997, IEEE Standards Dept.

Nobuo Murota (M06) received the B.S. degree in mechanical engineering

from Kyoto University, Kyoto, Japan, in 1985, and the M.S. and Ph.D. degrees
in civil engineering from the University of California at Irvine in 2000 and 2003,
respectively.
Currently, he is a Senior Engineer with the Department of Seismic Isolation
Engineering, Bridgestone Corporation Japan, Yokohama. His area of specialization is the research and development of seismic isolation systems.

Maria Q. Feng received the Ph.D. degree in mechanical engineering from the
University of Tokyo, Tokyo, Japan, in 1992.
Currently, she is Professor of Civil Engineering and Leader of the Environmental and Civil Infrastructure Layer of the California Institute of Telecommunications and Information Technology, with the University of California at
Irvine (UCI). Her research interests are primarily in the fields of innovative
and interdisciplinary science and technology for earthquake and wind protection, sensors and health monitoring, and damage detection of civil infrastructure
systems.

Gee Yu Liu received the B.S. degree in mechanical engineering from the
National Chiao Tung University, Hsinchu, Taiwan, R.O.C., in 1987, the M.S.
degree in mechanical engineering from National Taiwan University (NTU),
Taipei, in 1989, and the Ph.D. degree in applied mechanics from NTU in 1996.
Currently, he is Associate Research Fellow with the National Center for Research on Earthquake Engineering (NCREE), Taipei, where he has been since
1996. His area of specialization is earthquake engineering with an emphasis on
the study of nonlinear behavior of structures and seismic performance of lifeline
systems.