1806 S.-H.

Lee: Application of High Voltage Current Limiting Fuse Model Using ATP-Draw

Application of High Voltage Current Limiting

Fuse Model Using ATP-Draw
Sei-Hyun Lee
Korea Polytechnic College 4, Dept. of Electrical
Measurement and Control Engineering, South
Korea

ABSTRACT
This paper presents the results of arcing energy by the initiation of arcing, the arc
clearing voltage, the system impedance (X/R), and the making angle. To perform this
analysis, current limiting fuse (CLF) model was developed using the ATP-Draw
program. The violent activation of arc is affected by the voltage level, the voltage rise
rate when the arcing starts, and the minimum and the maximum positions of the
arcing energy that can be moved to zero degree by the X/R, and the more severe
interrupted condition has to be considered. The model developed using ATP-Draw
enables a non-expert to use it without spending excessive time and effort.
Index Terms — Modeling, current limiting fuses, X/R, power factor, arcing energy,
fault current, prospective current.

1 INTRODUCTION interrupt the fault current and the cut-out switch needs a
minimum time of over a minimum of a half cycle. Hence,
THE prospective fault current caused by line-to-line or when the fault is generated faster, current interruption is
line-to-ground faults has increased significantly in recent years required to limit energy inflow. The CLFs are suitable for
because of the increase in power system capacity. This has interrupting in such cases. It can limit the prospective current
generated situations where fault current has exceeded the to less than a quarter cycle after the fault, and then clear the
interrupting rating of the protection system. fault totally within less than a half cycle. To develop the CLFs,
The current limiting fuses (CLF)'s current limiting action research investigating the characteristics of fuse operation
starts over a certain current called the threshold or let-through before testing has been done using computer simulation.
or cut-off current and substantially reduce the amount of In earlier research, several models of CLF were developed
energy and the excessive stresses on the power system because [9-14], most of them based on a mathematical representation
of current limiting, and improve power quality by supporting of the arc physics. These models include transient heating and
the system voltage during operation. Short-circuit tests are fusion of notched strip elements in sand, arc ignition, and
used to verify the performance of fuses during design and subsequent burn-back, radial expansion of the arc channels
development. These tests require the use of high-power-test- due to fusion of the sand, merging of adjacent arcs, and many
laboratory facilities, and are expensive and time-consuming. other second-order effects [15, 16].
Computer modeling of the operation of fuses under these
conditions can be used to study fuse characteristics, and in The simpler empirical model of CLF, developed for use
recent years, there have been significant improvements in the with electromagnetic transient program (EMTP) was carried
accuracy of the fuse models available [1- 4]. out as a function of source voltage making angle [10].
However, the dissipated energy during the operation can be
Nowadays, CLFs are used to protect various types of changed by system impedance (X/R) as well as prospective
equipments on an electric power distribution system. The current and making angle [2-4, 17-21].
protected equipment includes the following items:
transformer, feeders, and section of the system, capacitor, In this paper, modeling with respect to CLF is developed
motors, and circuits with motors on them, etc. [5-8]. more easily using ATP-Draw, which is a graphical mouse-
driven preprocessor for ATP, to allow estimation of the
To interrupt these fault currents, a circuit breaker, a cut-out interrupting performance of CLFs. It has been designed for the
switch and a CLF have been used in the distribution line. The specific purpose of helping research to improve the
circuit breaker needs a minimum time of four cycles to performance of CLF. To observe the influence of the
operating energy by the X/R, the prospective current, and the
making angle, the model based on the test results of actual
Manuscript received on 14 January 2010, in final form 2 September 2010. CLF operation has been carried out.

1070-9878/10/$25.00 © 2010 IEEE

IEEE Transactions on Dielectrics and Electrical Insulation Vol. 17, No. 6; December 2010 1807

2 CONSIDERATION OF OPERATION greater than before and the terminal voltage of the fuse rises.
When both the source voltage and the terminal voltage of fuse
A fuse is a part of the distribution line before a fault, but the
become equal, it means the slope of current equals to zero at
fault occurs when the fuse is operated by the short-circuit
that point and the current is limited.
current. The fuse element senses the fault current, melts, arcs,
and the arc plasma is absorbed in the filler. The fulgurite c
created by the arc rapidly increases the resistance of the fuse
element. Because of this resistance, the phase difference a
between current and voltage is small. If the fault is not cleared,
the body of fuse or the tank of the transformer can explode. d e
Therefore, before the fault energy reaches an explosive level,
b
it has to be cleared. CLFs limit the fault energy, I 2 rt more
effectively than conventional breakers or non-limiting fuses.

Figure 1. Interrupting circuit model for fuse.

In Figure 1, R is the line resistive load, L is the inductive load,

ω is the pulsation at 60 Hz, r is the non-linear resistance of fuse, I Figure 2. Waveform of fuse’s operating.
is the RMS prospective current, and θ is the circuit making angle.
For this circuit, the prospective current is given by a: VS source voltage
ωt
i = 2 I {sin( ωt + θ − ϕ ) − sin( θ − ϕ ) exp( )} (1) b: L
di inductive voltage
tan ϕ dt
Putting ωt = π in Equation (1). c: V f terminal voltage of fuse
2
The value of I t is the time integral of the square of the d: i current
instantaneous current passing through a fuse link, as show in
e: r non-linear resistance
the following equation (2).
t

∫
I 2 t = i 2 dt (2) And then as the current has negative slope, the equation
0 becomes as following.
The I 2 t multiplied by the resistance, unit constant through
which the current flows is equal to the energy(joules). di
V f = VS + L (5)
In Figure 2, before the fuse element is melted the dt
instantaneous current (d) and voltage are related by equation (3).
di VS − V f
V S = i( R + r ) + L
dt
(3) In this case: I=
L ∫ (6)

where the resistance r is non-linear with dependence on Subsequently, the fuse voltage exceeds the source voltage.
parameters of time, heating, etc. The resistance of the fuse As a result the slope of current becomes negative and then zero.
when compared with the line resistance can be ignored. The At this time, the terminal voltage is represented by equation (5).
slope of current after closing is positive and then increases The fuse voltage during arcing is higher than the source voltage
gradually. VS , because V f is the sum of the gradient term and V S . The

di current flows and then decreases, and at that time, if the slope is
V f = VS − L (4) constant, the inductive voltage becomes a constant. The fuse
dt voltage which is the sum of the source voltage and inductive
At t ≈ 0 ,the instantaneous terminal voltage of fuse, (c) is voltage changes with the phase difference becoming zero at the
nearly zero because the inductive voltage (b) is nearly the end. And then after L di is zero, the fuse voltage becomes the
same as the source voltage (a). However after the fuse element dt
melts, its resistance like curve (e), in Figure 2, becomes much same as the source voltage [17].
1808 S.-H. Lee: Application of High Voltage Current Limiting Fuse Model Using ATP-Draw

3 SAMPLE AND EXPERIMENTS voltage is controlled to a level below the specified switching
voltage to prevent damage to other equipments. Hence, the
In Figure 3, the design of CLF used is shown. The fuse
only method to develop the efficiency of the fuse is to keep
element is spirally wound on supporter. Its components are an
the fuse voltage lower than the maximum voltage, which can
outer tube made of glass fiber, a sand filler, an internal
damage other equipments, until the fault current falls to zero.
supporter, which is ceramic with alumina above 70% and a
twisted and pressed fuse element, which is made of silver.

Figure 3. Cross section diagram of the tested fuse.

The interrupting test equipment is shown in Figure 4.

Experiment was carried out under the test voltage of 21 kV Figure 5. Operation of a general CLF.
and the prospective current of 1.2 kA. This equipment can The fuse voltage waveform is dependent on the way in
control a making angle and series impedance. which the arc plasma resistance is controlled when the fault
current is interrupted. In this paper, to analyze the operation of
the CLF with the fuse terminal voltage varied by the arc
plasma, the characteristics during three periods are considered
as shown in Figure 5.
In Figure 5, the parameter for “A” period determines the
melting energy. The energy I 2 rt , without any heat sink, is
Gen synchronous generator
determined as the melting point. This energy is the input as a
MS making switch
constant value decided by experiment. At “B” period, there is
CLR current limiting reactor
a rise in voltage and time of the fuse terminal, until it reaches
Tr Transformer
the maximum voltage. The rise time of the terminal voltage
CT current transformer
can be changed by the circuit conditions. The rising voltage is
ACB auxiliary circuit breaker represented by the capacitor charging voltage. It takes some
RL load resistor time in the real fuse for the voltage to rise very sharply as the
LL load reactor state varies from the solid to the vapor. The start point of the
Fuse fuse under test triangular wave is not the current limiting point. It is, in fact,
the melting point when the voltage reaches the peak voltage
with the triangular wave. Although after the melting ceases
and the element starts to cut off subsequently, some small
current flows due to the remnant ionization. Hence, the
Figure 4. Circuit for interrupting test and experimental scene charging voltage rises continuously. “C” area is the unique
decreasing characteristic of the voltage until the fault current
4 CHARACTERISTIC PARAMETERS goes to zero. The rise time and voltage are calculated by
The main characteristic is to limit the fault current and the equation (7) [21]. In the “B” period, the voltage and the
limiting effect after melting is proportional to the fuse terminal decreasing current stage are simulated by using the
voltage. Thus, the higher the fuse voltage the faster is the nonresistance characteristic in ATP-Draw.
decrease in fault current. The ideal fuse operates when the
current is instantly interrupted when the arc forms. However, Vf =
∫ idt (7)
if the fault current is interrupted very rapidly, the slope of the c
fault current becomes very large and the inductive voltage can c : Equivalent capacitor
exceed the source voltage.
V f : Maximum voltage at the terminal
Such a high switching voltage can damage other equipment
items connected to the system. Therefore, the maximum fuse i : Current during the voltage rise
IEEE Transactions on Dielectrics and Electrical Insulation Vol. 17, No. 6; December 2010 1809

power
5 APPLICATION CONSIDERATIONS
5.1 COMPARISON OF ACTUAL TEST AND THE
SIMULATION
In Figure 7a, the voltage, the current, and the calculated
resistance waveforms from laboratory tests under International
Standard IEC60282-1 with the fuse's rating, 24 kV, 25 A and the
simulated waveforms are compared for the power source, 21 kV,
the prospective current, 1.2 kA, and the system frequency, 60 Hz.

Before After
Switch Condition
status status
SW1 Closed I 2t > melting Opening

SW2 Opened I 2t > melting Closing

arrived at V f in
SW3 Closed Opening
period B
arrived at V f
SW4 Opened Closing
in period B
SW5 Closed maximum V f Opening

SWn Opened maximum V f Closing

Figure 6. CLF model used in ATP-Draw.

The circuit of the modeled fuse is shown in Figure 6 and

its operation is explained as follows. When the main switch
is closed by MODEL function, which can detect the
magnitude of the source voltage, the prospective fault
current flows through R2 . The melting time can be Figure 7. Comparisons between the experimental result and the simulation of
CLF at ϕ = 6.5 °, (a) voltage and current wave, (b) instant resistance
determined from the Joule energy. The energy of melting is
calculated from the actual test.
a constant value; hence, the melting time can be varied by
the power source. The melting is unaffected by other The maximum switching voltage is 60.5 kV, and the
external influences. After the time required for the melting simulated voltage is 59.9 kV. Hence, the difference is less than
energy to be passed, the closed switch, SW1 is opened and, 1%. The tested cut-off current is 1.4 kA, and the simulated
at the same time, SW2 is closed. The initial rising value of current is 1.38 kA. Thus, the difference in this case is about
1.43%.
voltage at the fuse terminal is calculated by C 2 . If the fuse
Table 1. Comparison of experimental and simulation results.
voltage reaches the value that is already decided at C2 , the
closed switch SW3 is opened and SW4 is closed at the Voltage Cut-off Melting   Arcing Operating
2 2
same time. The voltage rise is closed to actual wave by a Max. current I t I t I 2t
series of similar processes. If the voltage reaches the value
Experiment 60.5kV 1400A 1652 A2s 2608 A2s 4260 A2s
that is already decided at C3 , the closed SW5 is opened and
SWn is closed; hence, the non-linear resistance R3 operates Simulation 59.9kV 1380A 1652 A2s 2534 A2s 4186 A2s
for simulation. The parameter values were decided by the Δ% 0.99 1.43 0.00 2.84 1.74
experimental waveforms.
1810 S.-H. Lee: Application of High Voltage Current Limiting Fuse Model Using ATP-Draw

Also, the differences of the melting (pre-arcing) energy, the the resistance is gradually larger, because a distribution line
arcing energy, and operating energy (melting energy + arcing is longer. As the X/R is varying, the time required for the
energy) are 0%, 2.84%, and 1.74%, respectively. The results melting or arcing of fuse elements can be differed from the
of experiment and simulation are in agreement within 3% current rise rate caused by the making angle. The high
deviation. As a result, we can see that the CLF model current rise rate decreases the difference between the
developed can be applied to analyze the CLF characteristics making angle and the starting angle of arcing. The
(see Table 1). activation of arc plasma depends on the magnitude of the
source voltage when the arcing starts. If the magnitude of
the voltage is higher and its rise rate is positive, the arc
5.2 COMPARISON OF VOLTAGE AND PROSPECTIVE
during fault can be more violent.
CURRENT BY MAKING ANGLE UNDER THE
DIFFERENT X/R CONDITIONS
5.3 TENDENCY OF ARCING ENERGY BY VARYING
MELTING ENERGY UNDER THE X/R AND MAKING
ANGLE CONDITION
To take an account for the rated current, which is decided
by fuse element’s thickness or material, etc. the melting
time decided as the melting energy is varied and one of the
results is shown in Figure 9. The melting energy, I 2 rt , of
the curves (a), (b) and (c) in Figure 9 is 1 per unit (pu), 7
pu, and 20 pu, respectively. The curves (a’), (b’), and (c’)
are current and the arcing starting point of the curves (a),
(b), and (c) at 74.4°, 106.5°, and 148°, respectively.
Making angle is 40°.
As the melting energy, I 2 rt , is varied, we can see that the
starting angle of arcing and the arcing I 2 rt can be varied.

(a) Asymmetrical prospective current with X/R, 9.4

Figure 9. Current and voltage wave by a variable of the melting energy of

(b) Asymmetrical prospective current with X/R, 0.5
CLF at ϕ = 40 °.
Figure 8. Prospective current shape for ϕ = 0 ∼ 1 ° referred to the voltage
under the different X/R conditions.
Figure 8 shows the voltage and current waveform for the When the arcing starts, the magnitude of source voltage and
X/R (9.4 and 0.5) and the symmetrical prospective current the voltage rise rate can make the arc activated. In Figure 10,
of the curve (a) and (b) is 1.1 kA and 9.4 kA, respectively. the simulation circuit condition is the X/R of 6.6, the
The X/R and the phase differentiation are larger near a prospective current of 1.6 kA, and making angle of 40°.
power station, i.e. the X/R ratio in medium voltage systems
is high, of the order of 16-80, the lower value applicable to Arcing energy obtained by varying the melting energy I 2 rt
smaller distribution system of 10-15 MVA and the larger during fault is shown in Figure 10. We can see that higher of
values applicable close to plant generators of the order of melting I 2 rt can cause higher arcing I 2 rt for making angle.
70-80 MVA. Hence, the current is zero when the fault can In case of the minimum arcing energy, the 1 pu is 3.8 kJ at
be generated around the peak of a system voltage, and the 140°, the 7 pu is 12.7 kJ at 120°, and the 20 pu is 17 kJ at 100°.
condition to interrupt fault can be more severe. However, In addition, in case of the maximum arcing energy, the 1 pu is
IEEE Transactions on Dielectrics and Electrical Insulation Vol. 17, No. 6; December 2010 1811

54.6 kJ at 20°, the 7 pu is 105 kJ at 20°, and the 20 pu is 180

kJ at 0 degree. The arcing energy, 7 pu is 1.9 times, 20 pu is
3.3 times on the basis of 1 pu

Making angle[ °]
(a) (b) (c)
Figure 11. Arcing energy of the CLF by the X/R with the different melting
energies at ϕ = 0 ∼ 180 °.

Consequently, we can observe that the arcing energy shows the

Figure 10. Arcing energy by a variation of the melting energy of CLF at tendency to increase with the increasing difference between the
ϕ = 40 °. making angle and the initiation angle of arcing. Furthermore, the
position of the maximum and minimum of the arcing energy is
The melting energy is assumed as a constant value. Hence, gradually shifted to 0° when the X/R is increased. Thus, the more
the arcing energy’s tendency during the fault is similar to the severe test condition of the CLF having varied rated current in the
operating energy. Thus, the minimum of arcing energy can be same test condition can be differed. Therefore, to test more severe
differed by the initial position of arcing in spite of the same conditions, the initiation position of arcing has to be considered.
making angle. In case of the 1 pu, the arcing energy is
minimized at 140°. However, in case of the 7 pu and 20 pu, 5.4 THE TENDENCY OF ARC ENERGY BY THE
the arcing energy is at 120° and 100°, respectively. Hence, the ABILITY OF ARC QUENCHING
point of the minimum arcing energy is shifted to the left axis.
Furthermore, the maximum arcing energy of the 1 pu and 7 pu
is shown at 20°. But in the case of 20 pu, it is shown at 0°. As
a result, the tendency of the maximum arcing energy is shifted
to 0°, as well. Thus, it can be concluded that the initiation
point of arcing is differed by the difference of the melting
energy and the activation of arc is affected by the voltage level,
the voltage rise rate at the position of arcing.
The arcing energy is shown at the making angle from 0
to 180° for the X/R 9.4, 6.6, 1.7, and 0.5, respectively, in
Figure 11. Figure 11a shows the arcing energy of 1 pu
where the maximum of arcing energy is 54.2 kJ when the
making angle is 20° and the X/R is 9.4. The maximum of
arcing energy is 49 kJ when the making angle is 40° and
the X/R is 0.5. In case of curve (b) showing the arcing
energy of 7 pu, the maximum of arcing energy is 104 kJ
when the making angle is 0° and the X/R is 9.4. The
maximum of arcing energy is 78 kJ when the making angle
is 60° and the X/R is 0.5. Current and voltage wave by a variation of the arc
Figure 12.
quenching of CLF at ϕ = 40 °.
In case of curve (c) with 20 pu, the maximum of arcing
energy is 161 kJ when the making angle is 0° and the X/R To determine the characteristics of the arc quenching, which
is 9.4. The maximum of arcing energy is 127 kJ when the depends on the form and number of notch, the material of
making angle is 60° and the X/R is 0.5. Consequently, we element, inserted filler, and others, the variation of the arcing
can observe that the arcing energy shows the tendency to energy by the arcing clearing voltage is shown in Figure 12. The
increase with the increasing difference between the making arcing clearing voltage in Figure 12 (A) is (a), (b), and (c), having
angle and the initiation angle of arcing. values of 1 pu, 2.4 pu, 2.7 pu, respectively, and the curves (a’),
1812 S.-H. Lee: Application of High Voltage Current Limiting Fuse Model Using ATP-Draw

(b’), and (c’) are the current. The curve (B) in Figure 12 is the maximum of arcing energy is 35.5 kJ when the making angle
instantaneous resistance during arcing. The rise rate of resistance is 40° and the X/R is 0.5.
shows a similar tendency, but the time of the position to be above
1000 ohms in curves (a’’), (b’’), and (c’’) is 7.37 ms, 7.97 ms,
and 8.7 ms, respectively. Initially, the resistance curve, (c’’), is
increased among these curves because the arcing clearing voltage
is the highest. The initiation of arcing is similar.
However, the arcing cleared time for 1 pu, 2.4 pu, and 2.7 pu,
is 160°, 172°, and 188°, respectively. Thus, we can conclude that
the arcing energy during the fault can be varied because of the
magnitude and the rise rate of source voltage when the arcing
starts. The arcing energy by making angle and initiation of arcing
is shown in Figure 13.

Making angle[ °]
(a) (b) (c)
Figure 14. Arcing energy of the CLF by the X/R at making
angle ϕ = 0 ∼ 180 °.

In case of curve (c) with the arcing energy of 2.7 pu, the
maximum of arcing energy is 41 kJ when the making angle is
40° and the X/R is 9.4. The maximum of arcing energy is 29.4
kJ when the making angle is 60° and the X/R is 0.5.
Consequently, we can see that the arcing energy shows the
tendency to decrease with the increasing arcing clearing voltage.
With the increasing arcing clearing voltage, the arcing energy is
observed to decrease, because the arcing clearing voltage is
related to the ability of arc quenching. The position of the
maximum and minimum of arcing energy is gradually shifted to
0° with the increase in X/R. Therefore, to increase the ability of
arc quenching, the terminal voltage of CLF should be higher
than the source voltage when the arc is initiated, and the voltage
Figure 13. Arcing energy by making angle and initiation of arcing. should remain unchanged until the arcing is cleared.

The making angle is 40° and 140°. The starting of arcing is 5 CONCLUSION
55° (circle) and 156° (square). The difference is about A fuse model based on experimental results has been
15∼16°. The voltage level is 23 kV when the arcing starts at developed. This model using ATP-Draw is more convenient to
55°. At this time, the voltage rise rate (dv/dt) is positive. When use and can be employed to calculate CLF performance for a
the arcing starts at 140°, the magnitude of voltage is 7.5 kV variety of fault conditions. In this paper, the agreement
and the voltage rise rate (dv/dt) is negative. The arcing energy, between the actual test results and the simulation results is
when the making angle is 40°, is increased by 143% when within 3%. Using this model, the arcing energy characteristics
compared with that when the making angle is 0°. However, of the CLF can be estimated by varying the melting energy,
the arcing energy when the making angle is 140° is decreased the arc clearing voltage, the X/R, and the making angle.
by 22% and its arc energy is minimum. As a result, the initiation of the arcing based on the physical
The starting angle of arcing is an important parameter to characteristics of CLF, although the making angle is the same,
decide the arcing energy. The arcing energy is shown at the can be differed. Furthermore, the activation of arc is affected
making angle from 0 to 180° for the X/R 9.4, 6.6, 1.7, and 0.5, by the voltage level and the voltage rise rate at the initiated
respectively, in Figure 14. Figure 14(a) shows the arcing position of arcing. Moreover, the minimum and maximum of
energy of 1 pu, where the maximum of arcing energy is 63.5 the arcing energy by the X/R is shifted to 0°. Hence, more
kJ when the making angle is 20° and the X/R is 9.4. severe interrupting condition can be moved.
The maximum of arcing energy is 62.9 kJ when the making The developed model can be a valuable tool for designers,
angle is 40° and the X/R is 0.5. In case of curve (b) with the as it can estimate the influence of the fuse operation
arcing energy of 2.4 pu, the maximum of arcing energy is 45 characteristics during fault using parameters and can reduce
kJ when the making angle is 40° and the X/R is 9.4. The spending time and effort for development.
IEEE Transactions on Dielectrics and Electrical Insulation Vol. 17, No. 6; December 2010 1813

REFERENCES [15] S. Memiaghe, W. Bussière and D. Rochette, “Numerical method for pre-
arcing times: Application in HBC fuses with heavy fault-currents”,
[1] Lj. Kojovic and S. Hassler, “Application of current limiting fuses in Conf. on Electric Fuses and Their Applications (ICEFA), pp. 127-132,
distribution systems for improved power quality and protection", IEEE 2007.
Trans. Power Del., Vol. 12, pp. 791-800, 1997.
[16] D. Rochette, W. Bussière, R. Touzani, S. Memiaghe, G. Velleaud1 and
[2] F. Muench, "Current limiting fuses: Application considerations and P. André, “Modelling of the pre-arcing period in HBC fuses including
cautions", IEEE Power Eng. Soc., Transmission & Distribution Conf., solid - liquid – vapour phase changes of the fuse element”, Conf. on
Vol. 3, pp. 1238-1240, 2003. Electric Fuses and Their Applications (ICEFA), pp. 87-93, 2007.
[3] A. Hirose, “Mathematical analysis of breaking performance of current- [17] T. Tanaka, H. Kawaguchi, T. Terao, A. Babasaki and M. Yamasaki,
limiting fuses”, Conf. Electric Fuses and Their Applications (ICEFA), “Modeling of fuses for DC power supply systems including arcing time
Liverpool, UK, 1976. analysis”, Telecommunications Energy Conf., INTELEC, pp. 135-141,
[4] R. Wilkins, “Generalised short-circuit characteristics for H.R.C. fuses”, 2007.
Proc. IEE, Vol. 122, pp. 1289-1294, 1975. [18] T. Tanaka, T. Tanaka, T. Babasaki and M. Mino, “Fuse blowing
[5] R. Wilkins, T. Saengsuwan and L. O’Shields, “Short-circuit tests on characteristics for HVDC power supply systems”, Telecommunications
current-limiting fuses: Modeling of the test circuit”, Proc. IEE, Vol. 140, Energy Conf., INTELEC, pp. 1-6, 2008.
pp. 30-36, 1993. [19] A. F. Howe and C. M. Jordan, “Skin and proximity effects in
[6] P.O. Leistad, H. Kongsjorden and J. Kulsetbs, “Simulation of short- semiconductor fuselinks”, 4th Intern. Sympos. Switching Arc
circuit testing of high-voltage fuses”, Conf. on Electric Fuses and Their Phenomena, Technical University of Lodz, Poland, pp. 324-327, 1981.
Applications (ICEFA), pp. 220-226, 1984. [20] R. Wilkins, H. C. Cline and E. J. Knapp, “Fuse protection of IGBT
inverters”, Intern. Conf. on Power Conversion and Intelligent Motion
[7] S. Gnanalingam and R. Willuns, “Digital simulation of fuse breaking
(PCIM'98), pp. 370-378, 1978.
tests”, Proc. IEE, Vo1. 127, pp. 434-440, 1980.
[21] Y. Kawase, T. Miyatake, and S. Ito, “Heat Analysis of a Fuse for
[8] A. Wright and K.J. Beaumont, “Analysis of high-breaking-capacity fuse Semiconductor Devices Protection Using 3-D Finite Element Method”,
link arcing phenomena”, Proc. IEE, Vo1. 123, pp. 252-258, 1976. IEEE Trans. Magnetics, Vol. 36, pp. 1377-1380, 2000.
[9] J. Douglas, "Solving problems of power quality", EPRl Journal, 1993.
[10] M. Dolegowski, “Calculation of the course of the current and voltage of
a current-limiting fuse”, Conf. Electric Fuses and Their Applications
(ICEFA), pp. 218, 1976.
[11] A. Petit, G. St-Jean and G. Fecteau, "Empirical model of a current- Sei Hyun Lee was born in Daejeon, Korea, in 1967. He
limiting fuse using EMTP", IEEE Trans. Power Delivery, Vol. 4, pp. received the B.Sc., M.Sc. and Ph.D. degrees in electrical
335-341, 1989. engineering from the Chungnam National University,
[12] S. Gnanalingam and R. Wilkins, "Digital simulation of fuse breaking Korea, in 1992, 1994 and 1998, respectively. He joined
tests", Proc. IEE, Vol. 127, pp. 434-440, 1980. the Korea Polytechnic College 4, where he has been an
associate professor of electrical measurement and control
[13] O. Bottauscio et al., "Influence of short-circuit power factor on the cut-
engineering since 1997. He was with the University of
off current of low voltage fuses", Conf. Electric Fuses and Their New South Wales (UNSW), as a post-doctorial fellow from 2003 to 2004.
Applications (ICEFA), pp. 201-205, 1991.
His main research interests are surge suppressors, lightning arresters, high
[14] V.N. Narancic and G. Fecteau, "Arc energy and critical tests for HV voltage power fuses and switchgears.
current limiting fuses", Conf. Electric Fuses and Their Applications
(ICEFA), pp. 236-251, 1984.