2502 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 29, NO.

6, DECEMBER 2014

Switching Overvoltage Measurements

and Simulations—Part I: Field Test
Overvoltage Measurements
Juan. A. Martinez, Senior Member, IEEE, Dan Goldsworthy, Senior Member, IEEE, and
Randy Horton, Senior Member, IEEE

Abstract—This paper presents the work carried out by the the transient is initiated and the trapped charge on the phases).
IEEE Working Grou[ on field-measured overvoltages and their Therefore, a number of simulations using the same system have
analysis for validating power system component models to be to be made with the time of energization modified in each sim-
used in switching transients studies. The work uses measurement
data obtained from field tests performed by Bonneville Power ulation either in a predictable manner (i.e., for determining the
Administration in June 1995 where switching overvoltages were peak overvoltage) or statistically (for obtaining an overvoltage
measured on one of its 230 kV lines. This paper includes a descrip- probability distribution) [1]–[6]. Thus, model validation must
tion of the switching tests and the main results derived from field be performed using specified parameters (e.g., point on wave in
measurements. Details of the switching procedures that were fol- which the transient is initiated, etc.).
lowed to calibrate component models used to match field-recorded
waveforms can be found in the Part II paper. Field measurements are the preferred method of validation
of models for switching transients. This is due to the range of
Index Terms—Power system switching transients, switching
frequencies associated with most switching transients and the
transients, transmission system.
fact that the initiation of the transient can be predefined; that is,
there is no randomness involved in the origin of the transient
I. INTRODUCTION events. However, field measurement data obtained with accu-
rate measuring equipment are relatively rare. Some field mea-

S WITCHING transients in power systems are caused by the

operation of breakers and switches [1]–[11]. The switching
operations can be classified into two categories: 1) energization,
surement data have been presented to date for validation of com-
puter models. Reference [12] presents some cases with a good
agreement between simulation results obtained with an Electro-
including reclosing, and 2) de-energization. The former cate- magnetic Transients Program (EMTP)-type program and either
gory includes energization of lines, cables, transformers, reac- field measurements or transient network analyzer (TNA) results.
tors, or capacitor banks. The latter category includes current in- This paper summarizes some of the work carried out by the
terruption under faulted or unfaulted conditions. IEEE Working Group on Field Measured Overvoltages and
The results from the study of switching transients are useful Their Analysis. A primary goal of the working group is to quan-
to 1) determine overvoltage stresses on equipment; 2) select ar- tify the capability of transient programs to accurately predict
rester characteristics; 3) calculate the transient recovery voltage switching overvoltages, particularly on transmission lines which
across circuit-breaker contacts; 4) analyze the effectiveness of do not utilize mitigation measures to control switching surge
transient mitigating devices (e.g., preinsertion impedance, con- overvoltages (e.g., 230 kV class systems). Validating transmis-
trolled closing); and 5) determine overvoltage factors for live- sion-level power component models and simulation methods
line maintenance. requires accurate measurements of actual switching transients.
The level of detail required in the model varies with the study. The model validation work takes advantage of the switching
In addition, the results are highly sensitive to the value of cer- surge tests performed by Bonneville Power Administration
tain parameters (e.g., the point on the voltage wave with which (BPA) in June 1995 on the Big Eddy–Chemawa 230 kV line.
BPA carried out extensive single- and three-phase switching
Manuscript received September 15, 2013; revised December 20, 2013; ac- tests on this line, with and without trapped charge, from which
cepted January 25, 2014. Date of publication March 04, 2014; date of current a significant amount of information was recorded [13].
version November 18, 2014. Paper no. TPWRD-01059-2013.
The authors are with the Working Group on Field Measured Overvoltages This Part I paper provides a description of the BPA field test,
and their Analysis, General Systems Subcommittee, Transmission and Distri- including the purpose, procedures, and measurements, along with
bution Committee (e-mail: [email protected]; [email protected]; a summary of the main results. The Part II paper details simulation
[email protected]).
The Working Group on Field Measured Overvoltages and their Analysis is work carried out to validate the models applied in this study.
with the General Systems Subcommittee, T&D Committee, IEEE Power & En-
ergy Society. Task Force Members (in alphabetic order): Dan Goldsworthy, II. TEST LINE AND INSTRUMENTATION
Randy Horton, Albert Keri, Ilhan Kocar, Jean Mahseredjian, Juan A. Martinez,
Ben Shperling, and Tom Tobin. A. Test Line
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. The Big Eddy–Chemawa 230 kV line is a typical, long,
Digital Object Identifier 10.1109/TPWRD.2014.2303575 high-voltage (HV) line without switching surge overvoltage

0885-8977 © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
MARTINEZ et al.: SWITCHING OVERVOLTAGE MEASUREMENTS AND SIMULATIONS—PART I: FIELD TEST OVERVOLTAGE MEASUREMENTS 2503

control, such as surge arresters, breakers with closing resistors

or controlled closing. It is 187 km (116.4 mi) long and uses a
single conductor per phase. Approximately one-third of the line
is double-circuit construction and two thirds is single-circuit
construction. Details of the line configuration and surrounding
230 kV system can be found in the Part II paper. There are no
surge arresters on this line—only station entrance rod gaps at
each end, which will spark over for overvoltages and create
a fault. There are also no transformers or magnetic voltage
transformers to drain away trapped charge.
The line is switched with two SF circuit breakers of different Fig. 1. One-line diagram of the test line and measurements of line voltages,
manufacturers using single mechanisms and no closing resis- bus voltages, and line currents.
tors. When a fault occurs on the line, the breakers will trip and
then high-speed reclose after an open time of approximately 500
the transients associated with the bus voltages would not be sub-
ms (30 cycles). The high-speed reclose can result in substan-
stantial. The line currents were obtained by using the line-side
tial overvoltages on this line because of 1) the trapped charge
circuit-breaker bushing current transformers (CTs). Thus, some
on the unfaulted phases; 2) the near simultaneous closing of
reduced frequency response was introduced into the bus voltage
the three circuit-breaker poles; and 3) no overvoltage mitiga-
and line current measurements through the use of standard high-
tion equipment.
voltage devices.
At one end of the line is the Big Eddy 230 kV bus, a very
3) Data Acquisition: BPA recorded the data from each test
strong source with a maximum three-phase short circuit cur-
in digital format with a 1 MHz sampling rate using fiber-optic
rent of 46 kA (at the time the measurements were taken) and
data links to the voltage and current sensors.
numerous connected lines and equipment. In contrast, the
Chemawa end of the line is a relatively weak source, with few III. FIELD TEST MEASUREMENTS
lines and about 12 kA (at the time the measurements were
taken) of short circuit current availability. A. Line Switching Tests
The field tests carried out by BPA in 1995 included trans-
B. Background former switching and line switching. The line switching tests,
Rod gaps are installed at the ends of a transmission line to which are discussed in this paper, can be classified into two
protect substation equipment by sparking over during lightning- groups as described below [13].
caused overvoltages. These rod gaps are expected to rarely op- 1) Single-Phase Line Switching: Single-phase energization
erate during line switching, except under worst-case conditions. of the line eliminates the additional transients and coupling that
In November 1994, a rod gap sparkover occurred on the Big occur when other phases are energized. This provides a means
Eddy–Chemawa 230 kV line at the Big Eddy end during a high- of separating the direct transients from the coupled transients
speed reclose of the Chemawa breaker following a fault. The and reduces the number of variables when comparing measured
sparkover was unexpected and led to some minor equipment and simulated waveforms. The main purpose of these tests was
damage [13]. The main concern about the Big Eddy–Chemawa to acquire switching surge waveforms for validation of transient
rod gap sparkover was that reclosing overvoltages, particularly models used in line switching simulations. To perform the tests,
with SF breakers, might be higher than expected. As part of the a disconnect switch was blocked open on two phases and the cir-
investigation into this event, a field test was performed to mea- cuit breaker was operated normally. These tests were performed
sure overvoltage levels that can occur on long transmission lines by energizing the line from each end with and without trapped
during high-speed reclosing. The findings could affect transmis- charge. To create the trapped charge, the breaker that was ini-
sion line and substation maintenance at BPA, such as clearance tially energizing the line was opened and reclosed in 500 ms (30
practices and minimum approach distances. cycles).
2) Three-Phase Line Switching: The highest switching over-
C. Instrumentation voltages occur while reclosing into a line with trapped charge.
Fig. 1 provides a simplified one-line diagram of the line and The three-phase switching tests were performed from each end
test instrumentation used to conduct the field test. of the line. Three-phase trip and reclose tests were performed
1) Line Voltages: The most critical results of the field test since they approximate high-speed reclosing of a faulted line.
were the line voltage measurements. Typical power system The tests were performed to acquire waveforms for validation
voltage measuring devices, such as CVTs and MVTs, do not of transient models and to provide statistical data on actual over-
provide the high-frequency or dc response needed to accurately voltages that could be expected during a high-speed reclosing
measure reclosing voltages or properly verify transient simu- event. Energizing the line from Big Eddy provided informa-
lations. On each phase at each end of the line, BPA installed tion about line switching from a strong source, while energizing
special R-C-R voltage dividers with a flat frequency response from Chemawa provided information about line switching from
from dc to 1 MHz. a relatively weak source. A 60 Mvar, 230 kV shunt capacitor
2) Bus Voltages and Line Currents: For the bus voltage at bank was in service at the Chemawa bus during the switching
each end of the line, the substation bus MVTs were used, since from the Chemawa end.
2504 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 29, NO. 6, DECEMBER 2014

Fig. 2. Single-phase switching test from Big Eddy with trapped charge (Test 1–04).

For each test, the circuit breaker initially energizing the line 341 kV (bus line) and the resulting overvoltage measured at
was opened for 500 ms (30 cycles) and then reclosed 3-phase. Chemawa was 500 kV. The traveling waves on B-ph, interacting
The opening was controlled to leave the same trapped charge between the line and the system at Big Eddy, took approximately
on the line for each test. During the field test, the average time two cycles to dampen out. The waveforms show the relatively
constant of trapped charge decay was measured and found to be complicated voltages induced on A and C phases compared to
approximately 1 min. The trapped charge voltage thus decreased the relatively simple waves on B phase.
about 1% during the 500 ms open time. This decay time constant Approximately 25 switching tests were performed from each
for trapped charge is consistent with other BPA field test mea- line end, involving a three-phase reclose into a trapped charge.
surements made on other lines with these voltage dividers. The Nearly all tests conducted from a particular end of the line had
3-phase reclosing test was repeated 20 times from each end, with the same trapped charge, within a few kilovolts. Table I pro-
the closing signal incremented by 18 electrical degrees for each vides a summary of the average trapped charge voltages and
test. Incrementing the electrical angle in this way provided a the highest overvoltages measured on each phase during the
uniform distribution across a 60 Hz cycle. About five additional three-phase switching tests. The trapped charges were approx-
tests were performed from each end with the breaker timing fo- imately 1.2 p.u. on the highest phase and 0.9 to 1.0 p.u. on the
cused around the closing times that generated the highest over- other two phases. The highest overvoltages ranged from 2.9 to
voltages. 3.3 p.u.
While switching from Big Eddy (a relatively strong source),
B. Trapped Charge and Overvoltages C-phase had the highest maximum and average overvoltages,
Fig. 2 shows a 40 ms window from a single-phase switching even though A-phase had the highest level of trapped charge
test with trapped charge. In this test, the Big Eddy breaker was (see Table I). While switching from Chemawa (a relatively
tripped and reclosed in about 500 ms. Phases A and C of the weak source), A-phase had the highest level of trapped charge
line disconnect had been blocked open so only B-ph of the line and the highest maximum and average overvoltages. Switching
was directly energized. The trapped charge voltage on B-ph of from either end produced overvoltages exceeding 3.0 p.u.,
176 kV is shown on the left side of the waveforms prior to with the highest at 3.30 p.u. The upper 25% of overvoltages
the breaker close. The A and C phase voltage waveforms show measured when switching from Big Eddy were higher than
a partial trapped charge (approximately 32 kV each) that had those when switching from Chemawa. This was expected
been induced from B-ph. The step voltage applied to B-ph was since a stronger source, with more connected lines (and lower
MARTINEZ et al.: SWITCHING OVERVOLTAGE MEASUREMENTS AND SIMULATIONS—PART I: FIELD TEST OVERVOLTAGE MEASUREMENTS 2505

TABLE I a distribution of the maximum measured overvoltages for the

(a) AVERAGE TRAPPED CHARGE ON LINE. (b) HIGHEST MEASURED sending and receiving ends of the line. This plot uses only the
OVERVOLTAGES
20 reclosing tests from each end that were timed to be equally
spaced across a cycle. Note from the graph that 25% of the tests
when switching from Big Eddy are 3.0 p.u. or greater and 20%
are above this level when switching from Chemawa. Chemawa
switching also produced larger percentages of overvoltages in
the range of 1.5 to 2.8 p.u.. Thus, overvoltage levels were similar
despite the significant differences in the short circuit strength
and complexity of the two buses. While switching from Big
Eddy, the sending-end voltages above 2.0 p.u. (approximately
70th percentile and above) were substantially higher than when
switching from Chemawa.
Three rod gap sparkovers occurred at the Chemawa end
during the tests while reclosing the line from Big Eddy. One
sparkover occurred on B phase with a voltage prior to the
sparkover of 560 kV (2.83 p.u.). Rod gap sparkovers occurred
on two phases in another test, where the voltage at which the
gap flashed was 542 kV (2.74 p.u.) on A phase and 532 kV (2.69
p.u.) on B phase. Although significantly higher overvoltages
were measured on these phases during other tests, the gaps did
not spark over, which demonstrates the effects of surge polarity
and the statistical nature of gap sparkover phenomena.
Fig. 4 shows the waveforms for all voltages and currents
measured during one of the three-phase reclosing tests from
Chemawa. This plot shows how the transients involved with the
reclose operation dampen out in about one cycle. The sending-
end line voltage and bus voltage are overlaid for each phase.
The trapped charge can be seen as the constant dc level of the
line voltage at the left side of the waveforms. The first breaker
prestrike occurred on A-ph where the line trapped charge and
the bus voltage were opposite polarity. The many complications
and waveform distortions on each trace are from traveling-wave
reflections, induced voltages from other phases, and multiple
breaker prestrikes. The travel time that a switching surge takes
to travel the length of the line can be seen to be about 0.6 ms.
The “round trip” travel time for a current pulse to return to the
sending end is about 1.26 ms. In this test, the A-ph voltage at
the Big Eddy (receiving) end reached about 3.0 p.u. 564 kV).

C. Multiple Prestrikes
The measurements taken during the three-phase reclosing
tests revealed an unexpected phenomenon—that breaker
closing operations into trapped charge can result in multiple
prestrikes [6], [14], [15]. The majority of breaker closings
Fig. 3. Distribution of maximum overvoltages measured during the test for
both receiving and sending ends. resulted in only a single prestrike; however, in a few tests, up
to four prestrikes occurred on one phase during a single closing
operation. During breaker prestrike, a current wave (initiated
source surge impedance) will produce a larger voltage step by arcing across contacts) travels down the line to the receiving
onto a switched line. However, in the range of 15%–75% of (open) end where it is reflected. As the reflected wave travels
overvoltages, this did not hold true and the levels were larger back toward the sending end of the line, it reduces the current
when switching from the Chemawa end, indicating additional to near zero along the line. When the reflected current wave
variables at work. reaches the sending end, it creates a current zero and allows
Along with the receiving-end overvoltages, the sending end the prestrike arc between the breaker contacts to extinguish,
of the line also experienced overvoltages which were higher isolating the line voltage from the bus voltage. The line voltage
than expected. These occurred because of induced voltages from may then increase due to travelling waves that continue to be
transients on other phases and because of the breakers having reflected from the receiving end and induced voltages. The
multiple prestrikes, which are discussed later. Fig. 3 provides voltage across the breaker then builds up until another prestrike
2506 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 29, NO. 6, DECEMBER 2014

Fig. 4. Three-phase high-speed reclose test while switching from the Chemawa end of the line (Test 5–71).

occurs. The next prestrike occurs at a lower breaker cross

voltage because the breaker contacts are then closer.
The multiple prestrike phenomenon can be seen on each
phase in Fig. 4, where the overlaid line and bus voltages deviate
from each other after the breaker current has been established.
At these points, the breaker current remains at zero until another
prestrike occurs. As shown in Fig. 4, the multiple prestrikes
add even more transients and distortion to already complicated
waveforms.
The recorded waveforms of voltages and currents were an-
alyzed and prestrike data were tabulated; namely, the voltage
as well as the time the prestrike occurred (relative to the ini-
tial closing time). From the tabulated data, the prestrike volt-
ages versus relative prestrike times were plotted. As an example,
Fig. 5 shows the plot of the Chemawa breaker prestrike data for Fig. 5. Chemawa breaker prestrike data for all phases.
all phases derived from reclosing operations [13]. Linear regres-
sion was used to determine the approximate voltage versus time
characteristic for each phase of both breakers. Table II lists the The Big Eddy breaker was independent pole construction
slopes derived from that analysis. (i.e., constructed with each phase in its own tank), and the max-
MARTINEZ et al.: SWITCHING OVERVOLTAGE MEASUREMENTS AND SIMULATIONS—PART I: FIELD TEST OVERVOLTAGE MEASUREMENTS 2507

Fig. 6. Identification of prestrikes on the Chemawa breaker during a three-phase reclosing operation.

TABLE II the number of prestrikes for this test is different for each phase
BREAKER DIELECTRIC SLOPES DURING CLOSING OPERATIONS and results in up to three on A phase [13].

IV. MODELING DATA FOR TESTED 230 KV LINE

The Big Eddy–Chemawa line can be divided into four main
sections with about 1/3 of the line as a double circuit and the
remainder as single circuit. Fig. 7 provides the basic line con-
ductor and configuration data necessary for line modeling. Fig. 8
provides the parameters for simplified partial source equivalents
imum closing time difference between the phases (or pole span) for each line end.
was approximately 3.7 ms. This pole span was determined by
taking the difference in time that the prestrike characteristic V. LABORATORY TESTS OF LINE ENTRANCE GAPS
of each phase crosses zero voltage (i.e., the point where the As part of the investigation into the original rod gap flashover,
breaker contacts are connected metal to metal). The Chemawa switching surge and corona onset tests were performed at the
breaker was constructed with all three phases in a single tank, BPA Laboratories on a standard BPA 230 kV transmission line
and the maximum pole span was 0.24 ms, which would indi- entrance gaps. The tested rod gap was similar to those on the Big
cate that all three phases close nearly simultaneously [13]. For Eddy–Chemawa line and set to the BPA standard 101.6 cm (40
both breakers, the time differences of the prestrikes among the in). Fig. 9 provides a photo of typical BPA 230 kV “shepherd’s
phases vary considerably, however, because of the differences in hook” rod gaps.
breaker cross voltage during closing, particularly with trapped Table III provides the results of the switching surge portion
charge. of the testing, including both measured flashover voltages and
Fig. 6 identifies the many prestrikes that occurred on the also values corrected to standard atmospheric conditions. Other
Chemawa breaker during the reclose test of Fig. 4. Note that weather conditions could lower or raise these numbers. As
2508 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 29, NO. 6, DECEMBER 2014

Fig. 7. Conductor cross sections for Big Eddy–Chemawa line.

TABLE III
LINE ENTRANCE GAP SWITCHING IMPULSE TEST RESULTS VOLTAGES IN PU
WHERE 1 p.u. 197 kV PK [16]

Fig. 8. Equivalent source parameters for each line end.

Notice that the minimum flashover voltage in per unit is 3.1

for positive and 3.2 for negative polarity overvoltages, and that
these values were exceeded during field tests (Table I).

VI. SUMMARY AND CONCLUSION

A series of single-phase and three-phase line switching tests
were performed on a 116-mi, 230 kV line. These tests were con-
ducted to both obtain switching surge data for transient simula-
tion model validation and to measure the approximate magni-
tudes of overvoltages that result from high-speed, three-phase
reclosing with a trapped charge on the line. The complexity and
short circuit strength of the sources at each end of the line varied
considerably, which offered a unique opportunity to evaluate the
effects that such complexities have on the resulting switching
transients. The circuit breakers at each end of the line were also
Fig. 9. BPA 230-kV line entrance gaps. different in their electrical characteristics during closing.
The critical line voltage measurements for this test included
high-quality voltage dividers to provide accurate readings and
Table III shows, flashover voltages varied more for positive a 1 MHz sampling rate to provide good frequency response al-
polarity surges than for negative. The rod gap critical flashover lowing for direct comparison with simulation results. The bus
voltage was found to increase with higher relative humidity voltage and line current measurements used existing substation
but not be sensitive to the waveform time-to-crest, which was MVTs and CTs, which did not provide quite as good of fre-
varied between 100 and 1000 . quency response but was adequate for the application.
MARTINEZ et al.: SWITCHING OVERVOLTAGE MEASUREMENTS AND SIMULATIONS—PART I: FIELD TEST OVERVOLTAGE MEASUREMENTS 2509

The high-speed, three-phase reclosure tests were performed REFERENCES

by tripping a breaker that was charging the line and reclosing [1] IEEE Guide for the Application of Insulation Coordination, IEEE Stan-
it 500 ms later into the trapped charge, similar to a high-speed dard 1313.2–1999, 1999.
reclose after a fault. The trapped charge levels were approxi- [2] Insulation Co-ordination, Part 2: Application Guide, IEC 60071–2,
1996.
mately 1.2 p.u. on the highest phase and slightly less than 1.0 [3] M. Erche, “Switching surges,” in Chapter 3 in Surges in High-Voltage
p.u. on the other phases. For each test the breaker was controlled Networks, K. Ragaller, Ed. New York, USA: Plenum, 1980.
to trip at the same electrical angle, leaving approximately the [4] A. Greenwood, Electrical Transients in Power Systems, 2 ed. New
same trapped charge for each test. The breaker was also con- York, USA: Wiley, 1991.
[5] A. R. Hileman, Insulation Coordination for Power Systems. New
trolled to receive a close signal at times that were uniformly York, USA: Marcel Dekker, 1999.
spaced across a 60 Hz cycle. [6] J. A. Martinez-Velasco, Power System Transients. Parameter Determi-
About 25 three-phase reclosing tests were performed from nation. Boca Raton, FL, USA: CRC, 2009.
[7] A. R. Hileman, P. R. Leblanc, and G. W. Brown, “Estimating the
each line end. At each line end approximately 20% of the tests switching-surge performance of transmission lines,” IEEE Trans.
resulted in an overvoltage of 3 p.u. or greater. The maximum Power App. Syst., vol. PAS-89, no. 7, pp. 1455–1469, Sep. 1970.
overvoltage measured was 3.3 p.u. while switching from the [8] CIGRE Working Group 13.05, “The calculation of switching surges —
Part I,” , Electra, no. 19, pp. 67–122, 1971.
strong source end. The tests confirmed that overvoltages at [9] CIGRE Working Group 13.02, “Switching overvoltages in EHV and
the receiving end of a line reach magnitudes that can result in UHV systems with special reference to closing and reclosing of trans-
rod-gap sparkovers, which actually occurred on two of the tests. mission lines,” Electra, no. 30, pp. 70–122, 1973.
[10] CIGRE Working Group 13.05, “The calculation of switching surges II.
The tests also captured the interesting phenomena of mul-
network representation for energization and re-energization studies on
tiple prestrikes by a breaker closing into a trapped charge. This lines fed by an inductive source,” Electra, no. 32, pp. 17–42, 1974.
is caused by the prestrike arc extinguishing, and results in the [11] CIGRE Working Group 13.05, “The calculation of switching surges III.
voltage at the sending end of a line reaching values that are Transmission line representation for energization and re-energization
studies with complex feeding networks,” , Electra, no. 62, pp. 45–78,
higher than were previously expected. Multiple prestrikes also 1979.
add additional complexity to the switching surge waveforms. [12] H. W. Dommel, A. Yan, R. J. Ortiz de Marcano, and A. B. Miliani,
The subject line switching tests provided measurements that Case Studies for Electromagnetic Transients, 2 ed. Vancouver, BC,
Canada: Univ, British Columbia, 1993.
can be used to accurately model transmission system compo- [13] S. M. Lowder, J. L. Randall, and D. Goldsworthy, “Big
nents for use in transient simulations when analyzing switching Eddy–Chemawa 230-kV line switching and transformer saturation
surges, and data that can be used for modeling breaker pre- tests,” Bonneville Power Administration, Portland, OR, USA, Rep.
TTL(E) 96-11, Jun. 7–8, 1995.
strikes. Typical characteristics of the dielectric strength across [14] D. A. Woodford and L. M. Wedepohl, “Impact of circuit breaker pre-
the breaker contacts, together with multiple prestrike models strike on transmission line energization transients,” presented at the Int.
can now also be developed and used for statistical switching Conf. Power Syst. Transients, Seattle, WA, USA, Jun. 1997.
surge studies. Improved modeling techniques based on the mea- [15] EPRI Report EL-4651, EMTP Workbook II 1987.
[16] D. Bradley, “Electrical tests on 230-, 115-, and 69-Kv line entrance
surement data provided herein is the subject of Part II of this gaps, Bonneville Power Administration, Portland, OR, USA, Rep.
paper. TTL(E) 95-224, Nov. 9, 1995.