Está en la página 1de 6

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 33, NO.

2, MARCH/APRIL 1997

383

Transient Recovery Voltage Considerations in the


Application of Medium-Voltage Circuit Breakers
David L. Swindler, Senior Member, IEEE, Paul Schwartz, Member, IEEE,
Paul S. Hamer, Senior Member, IEEE, and Stephen R. Lambert, Fellow, IEEE
AbstractMedium-voltage circuit breakers can fail to interrupt
three-phase fault currents when power systems have transient
recovery voltage (TRV) characteristics which exceed the rating of
the circuit breaker. This paper examines the application of 13.8kV generation and load switchgear for an oil refinery in which
circuit parameters as originally designed would have exceeded the
13.8-kV circuit breakers TRV ratings had corrective measures
not been taken. This paper illustrates this case and discusses the
basis of TRV, how TRV is assessed, and alternative actions taken
to bring circuits to within the 13.8-kV circuit breaker ratings.
Index Terms EMTP, medium voltage circuit breaker, transient recovery voltage.
Fig. 1. Simplified TRV circuit representation.

I. INTRODUCTION

uring the short-circuit current interruption process, at


current zero when current is interrupted, the system oscillates in accordance with its natural frequency. This systemgenerated voltage, called the transient recovery voltage (TRV),
is impressed across the opening breaker contacts and stresses
the gap insulation. The interrupting media following current
extinction is attempting to return from a state of good conduction to one having the attributes of a good insulator.
Thus, a race takes placethe interrupting media becomes
a good insulator while the system is applying increasing
TRV to the gap in an attempt to reignite or restrike the
arc. If the insulation recovers more quickly than the TRV,
then a successful interruption occurs. If not, then the arc is
reestablished, another 1/2 cycle or loop of current occurs,
and the interruption process is again attempted. This process
continues until successful interruption occurs (or until the
breaker fails).
Current reignition occurs because the interrupting medium,
during the first few microseconds following current interruption, is not yet a good insulatorit may still have a relatively
high conductance. If the TRV has a high rate of rise in
those first few microseconds, it may reestablish a current flow
sufficient to heat the arc column and restore conduction, If
Paper PID 9619, approved by the Petroleum & Chemical Industry
Committe of the IEEE Industry Applications Society for presentation at
the IEEE/IAS 1995 Pulp & Paper Industry Committee Conference and the
1995 Petroleum & Chemical Industry Technical Committee Conference,
Denver, CO, September 1014, 1995. Manuscript released for publication
August 1, 1996.
D. L. Swindler is with Square D Company, Smyrna, TN 37167 USA.
P. Schwartz is with Fluor Daniel, Inc., Irvine, CA 92730 USA.
P. S. Hamer is with Chevron Research & Technology Co., Richmond, CA
948020267 USA.
S. R. Lambert is with Power Technologies, Inc., Schenectady, NY
123011058 USA.
Publisher Item Identifier S 0093-9994(97)01542-9.

I R heating exceeds the breakers capability to remove heat


and cool the arc channel, then reignition occurs.
Restrikes are really dielectric breakdowns and can occur
anytime during the TRV cycle, although usually they are
associated with breakdowns later in the cycle (say tens to
hundreds of microseconds) when the TRV has sufficiently high
magnitude. Such an event is akin to a lightning flash-over
across an insulator.
It should be clearly understood that while the magnitude of
the short circuit current is important to successful interruption
(i.e., it must be less than the breakers rating), failure to
interrupt due to excessive TRV can occur with low-magnitude
short circuit currents as well as currents at or near full rating.
[1]
The simplified TRV circuit representation shown in Fig. 1
represents one phase of a three-phase power system with a
source voltage, transformer, transformer winding capacitance,
secondary circuit breaker and a three-phase ungrounded fault.
The shunt capacitance represents the total of the phase-toground capacitances of the transformer winding and that of
the secondary cables.
For a three-phase system involving a three-phase ungrounded fault, the TRV (V ) across the breaker contacts of
the first pole to clear as a function of time is
(1)
where

fault
time
total
total

;
(rad/s);
current magnitude (kA rms);
(s);
leakage inductance (henry);
leakage capacitance (farad).

00939994/97$10.00 1997 IEEE

384

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 33, NO. 2, MARCH/APRIL 1997

Fig. 2. Current interruption process.


Fig. 3. TRV (one-minus-cosine) curve.

Equation (1) is referred to as the one-minus-cosine TRV


characteristic wave shape. The constant term E in the equation
results from the assumption that the 60-Hz supply voltage is
practically a constant over the time period of the interest for the
TRV. On the other hand, the cosine term is the high-frequency
oscillatory component of the TRV which is actually a damped
function when the resistance is considered. The one-minuscosine response given by (1), although an approximation since
damping effect is ignored, clearly illustrates two important
TRV parameters, namely, the peak value of the TRV, designated as E and the time after current zero to reach the peak
value, designated as T . These two parameters are used in the
ANSI Standards to specify TRV-related circuit breaker ratings.
The one-minus-cosine TRV characteristic generally occurs in
circuit breakers at the secondary side of a transformer.
It can be noted from the preceding that TRV analysis
involves solving differential equations in the time domain.
Even for the simple system of Fig. 1, a manual solution is
complicated and tedious if the effects of L, C, and R are
all considered. A computer program that allows an electrical
system to be modeled and analyzed in the time domain is
referred to as the electromagnetic transients program (EMTP).
With this program, all capacitances in the system can be
represented either as discrete or distributed parameters together
with all inductances and resistances. The program solves the
system equations numerically in the time domain and the
values of the TRV are determined for specific time instants
within a user-specified simulation time.
Fig. 2 illustrates an oscillograph of short circuit current and
voltage versus time during an interruption. Notice the contacts
have separated at some significant current along the current
wave form. The current is carried to zero by the arcing medium
within the interrupter. At current zero, the supply voltage is
near a maximum.
II. CIRCUIT BREAKER TRV RATINGS
ANSI/IEEE C37.04 [2], ANSI C37.06 [3], and ANSI/IEEE
C37.09 [4] defines four different types of TRV conditions for
all types of circuit breakers. (The one-minus-cosine waveshape
applies for circuit breakers rated 72.5 kV and below. The other
waveshapes apply for breakers rated 121 kV and above.)
1) One-Minus-CosineWave shape which results from the
parallel combination of an inductor and capacitor and

commonly occurs when clearing a three-phase fault fed


through a transformer.
2) Exponential CosineWave shape for three-phase faults
supplied by both transformers and transmission lines.
3) Short Line FaultA saw-toothed wave shape for short
line transmission line faults.
4) Initial TRVAn additional ramp and plateau of voltage
added to the initial front of an exponential cosine wave
shape. This TRV is due to relatively close inductance
and capacitance associated with substation bus work.
For this case study, the one-minus-cosine TRV condition
was appropriate. The other TRV conditions are generally
associated with high-voltage transmission line application line
applications.
A. General-Purpose TRV Breaker Ratings
The one-minus-cosine TRV response is a time function
given by (1) and a plot of this response for a time period
corresponding to the first-half cycle of the oscillatory component of the TRV is shown in Fig. 3. As is typical of electric
power systems, the natural frequency of the TRV wave form
is considered much higher than the supply frequency of 60
Hz. Thus, it can be noted in Fig. 3 that the 60-Hz component
voltage of the TRV is constant or a straight line over the very
short time period of interest. The normal supply voltage is
considered to be at its peak value during this time period.
Two parameters, E and T , necessary to define the TRV
are clearly shown in this illustration. E represents the peak
value that the TRV may reach. In ANSI/IEEE standards for
breakers rated 72.5 kV and below, E is 1.88 times the rated
maximum voltage. T represents the time required to reach
the peak value. The value of T is a variable dependent upon
the voltage rating, type of circuit breaker, and the short circuit
current level. For 15 kV, indoor class circuit breakers at 100%
rated short-circuit current, the value of T is 75 s. Up to this
point in time, ANSI C37.06 Table I [3] has not shown a value
of T however, a value of 75 s for 15 kV has been generally
agreed upon by members of IEEE Switchgear Committee and
by NEMA. This value, along with other values of T for other
voltages, will be shown in the next revision of ANSI C37.06,
when approved.
As the level of short circuit current is reduced from its
maximum interrupting rating, the required TRV performance

SWINDLER et al.: TRANSIENT RECOVERY VOLTAGE CONSIDERATIONS

Fig. 4. General-purpose TRV breaker requirements.

of a breaker increases. The amount of increase in performance


beyond the rated TRV parameters is defined by ANSI C37.06
Table 6 [3].
As shown in Fig. 4, the severity of the TRV increases in
both the magnitude of E and in the shortening of the time of
T as the short circuit current is reduced.
The TRV rating is verified during circuit-breaker shortcircuit design testing in accordance with ANSI/IEEE C37.09
[4]. This is accomplished by evaluating the test circuit for
TRV characteristics prior to the short-circuit test. Capacitance
is removed or added to the test circuit to cause the value of
T to conform to or be less than the specified values. A circuit
having too little capacitance would cause a small value of T
and give rise to the possibility that the breaker would fail to
interrupt the intended magnitude of short-circuit current. A
circuit having too much capacitance would not meet the TRV
requirements and, although the circuit breaker may clear the
current, it would not be considered a valid short-circuit test.
B. Definite-Purpose TRV Breaker Ratings
Users have recognized that in some instances standard TRV
ratings are not adequate for all applications. This has given
rise to a concept of developing special circuit breakers having
greater TRV capability. The industry has proposed to call them
definite-purpose circuit breakers. This is not to be confused
with breakers that have additional capacity for back-to-back
capacitor switching performance. There have been several
proposals for definite-purpose TRV breakers.
Fig. 5 illustrates the latest proposal. This proposal suggests
that additional capacity be given to breakers operating at 7%
and 30% of interrupting rating. The TRV parameters at 60%
and 100% would remain unchanged.
Table I shows a comparison of E and T for the present
general-purpose TRV ratings and the proposed definitepurpose TRV ratings.
III. CASE STUDY: 13.8-kV CIRCUIT BREAKERS
A TRV study was made for the application of 15-kV class,
general-purpose indoor breakers using the EMTP. The EMTP
study one-line diagram study is shown in Fig. 6.
The study was dependent upon the identification of all
inductive and capacitive items within the circuit. Items like
potential metering transformers were important because of

385

Fig. 5. Definite-purpose TRV breaker proposed requirements.

TABLE

OF

GENERAL-

TABLE I
DEFINITE-PURPOSE TRV RATINGS

AND

their capacitance to ground. Items which tend to cause adverse TRV conditions are power transformers and short-circuit
current-limiting reactors.
The TRV study considered various operating and fault
conditions and an analysis was made for each configuration.
Since the TRV requirement is a function of the three-phase
fault current level, it was also calculated for each condition.
For each breaker performing an opening operation, the threephase short-circuit current level, peak TRV voltage (E ),
and time to peak (T ) were calculated. These calculations
were then compared to the TRV ratings of the breaker in
question. If the calculated value of T was equal to or higher
than the breaker rating, then the application was considered
appropriate. If the calculated value was lower than the breaker
rating, then the condition was addressed further. The parameter
E was not given significant consideration since the value T is
a more profound parameter when considering the performance
of the breaker.
Fourteen cases were simulated; the results are summarized
in Table II.
The calculated TRV (kV) vsersus time response curves were
generated for each case. The response curves for Case 1 and
Case 2 are shown in Figs. 7 and 8.
Case 1 represents a fault on Bus 1 to be interrupted by one
of the main breakers, 52-1. In this case, the TRV is generated
essentially by the transformer. As can be seen, this curve is
a typical damped one-minus-cosine wave form. The EMTP
program results indicated that for Case 1 the time T is in
the order of 12 s.
Case 2 is interesting because the TRV response at the
breaker is a composite of three TRV sources, both transformers
and the reactor. As can be seen in Fig. 8, the response has a
double frequency, one as a result of the transformers and the
other the result of the reactor. In this case, the first minor peak

386

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 33, NO. 2, MARCH/APRIL 1997

*Capacitors shown dotted were added as a result of this study to correct TRV.
Fig. 6. EMTP study one-line diagram.
SUMMARY

TABLE II
OF EMTP RESULTS

is given attention because it is responsible for a very high rate


of change in voltage near the current zero. These conditions
are more severe to some breakers than is the magnitude of the
final voltage. For the sake of simplicity, the first minor peak
is considered to define the value of T in spite of the fact
that it is not the ultimate maximum. This evaluation was not
extreme in this case since Case 1 had already displayed TRV
characteristics more severe than this approximation.
Prior to the generation of the data shown in Table II, it was
determined that the short-circuit current-limiting reactor was
producing excessive TRV for the bus tie breakers. A 20 000-pF
capacitor, in parallel with each phase of the reactor, was added
to correct for this problem. The balance of the evaluation as
shown in Table II, left problems being created only by the
transformers. Cases 12 and 14 are reruns of Cases 1 and 5
illustrating the correction of the problems by adding a 0.25F (to ground) surge capacitor per phase to the transformer
bank incoming bus, as shown in Fig. 6.

Analysis of circuit breaker performance, as shown in Table


III, compares the system-calculated TRV response to the
circuit breakers capability. General-purpose circuit breaker
performances which were exceeded by the circuit response
were highlighted with a bold box and bold type. Data shown
in Cases 1 and 5 are shown corrected in Cases 12 and 14. Case
2 also illustrates a problem. Since the source of TRV was well
identified and the nature of the case understood, it was deemed
necessary to rerun Case 2 with the 0.25- F capacitance added.
IV. CONCLUSION
The initial results for the 13.8-kV circuit breakers considered in this case study indicated that without modifications,
the TRV would have exceeded the breaker ratings.
Based upon this case study, the following recommendations
and guidelines can be used by the application engineer on
other medium-voltage configurations:

SWINDLER et al.: TRANSIENT RECOVERY VOLTAGE CONSIDERATIONS

ANALYSIS

OF

387

TABLE III
CIRCUIT BREAKER PERFORMANCE

Fig. 7. Case 1 TRV response.

1) TRV studies are recommended where the reliability of


the equipment is critical to plant operations, e.g., main
plant incoming distribution switchgear with normally
closed ties and high-impedance transformers.
2) TRV studies are recommended where large generators with generator breakers are installed. (Refer to
ANSI/IEEE C37.013 for specific recommendations.)
3) TRV studies are recommended where current-limiting
reactors are installed.
For the above applications, the resolution to cases where
the TRV exceeds the circuit breaker rating may be the simple
addition of surge capacitors.

Fig. 8. Case 2 TRV response.

If the user cannot justify performing an EMTP study, then a


conservative approach could be the addition of 0.25- F surge
capacitors installed on each phase on the incoming side of the
main circuit breakers.
If a current limiting reactor is applied, then capacitors
installed across the reactor phases should be discussed further
with the reactor manufacturer. If these parallel TRV suppression capacitors are not available, 0.25- F surge capacitors are
recommended on each phase of both terminals of the reactor. It
is likely that capacitors would be recommended in many cases.

388

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 33, NO. 2, MARCH/APRIL 1997

REFERENCES
[1] S. R. Lambert, Circuit breaker transient recovery voltage, in IEEE
Tutorial Course: The Application of Power Circuit Breakers, 93-EHO388-9-PWR.
[2] IEEE Standard Rating Structure for AC High-Voltage Circuit Breakers
Rated on a Symmetrical Current Basis, ANSI/IEEE Standard C37.041979 (Reaff 1988).
[3] AC High-Voltage Circuit Breakers Rated on a Symmetrical Current BasisPreferred Ratings and Related Required Capabilities, ANSI Standard C37.06-1987.
[4] IEEE Standard Test Procedure for AC High-Voltage Circuit Breakers
Rated on a Symmetrical Current Basis, ANSI/IEEE Standard C37.091979 (Reaff 1988).
[5] IEEE Application Guide for Transient Recovery Voltage for AC HighVoltage Circuit Breakers Rated on a Symmetrical Current Basis,
ANSI/IEEE Standard C37.011-1979.
[6] R. K. Smith, Tests show ability of vacuum circuit breaker to interrupt
fast transient recovery voltage rates of rise of transformer secondary
faults, 94 TD 026, T-PWRD. Jan. 1995, Westinghouse Science &
Technology Center, Pittsburgh, PA 15235 USA.
[7] W. G. Heinmiller, R. W. Katterhenry, S. R. Lambert, T. W. Stringer,
Transient recovery voltage failures of two 15kV indoor oilless circuit
breakers, IEEE T-PAS, pp. 2578-2584, Aug. 1982.

David L. Swindler (M58SM92) received the


B.S.E.E. degree from Case Institute of Technology,
Cleveland, OH, in 1958 with a Commission in the
U.S. Air Force.
He joined Square D Company in 1958 as a Field
Service Engineer. Shortly after joining with Square
D, he served four years in the Air Force as a
Chief of Communications and Electronics Officer.
He returned to Square D in 1962 where he served
10 years in the R&D Engineering Department of the
Heavy Industrial Control Division. In 1972, he was
appointed Chief Engineer of the Middletown, OH, plant of Square D, which
specialized in switchgear products. In 1984, he moved to the Smyrna, TN,
Switchgear Division, where he is now serving as a Senior Staff Engineer. He
has been involved in the development and testing of several switchgear-type
products and holds several design patents. He has served on various NEMA,
IEEE, and IEC Standards Development Committees, and he has authored
several technical papers.
Mr. Swindler is a Registered Electrical Engineer in the State of Tennessee.

Paul Schwartz (M75) received the B.S. degree


in engineering physics from the University of Colorado, Boulder, in 1975.
Since 1975, he has worked for Fluor Daniel,
Inc., Irvine, CA, and is presently a Chief Principal
Electrical Engineer in Power Generation. He is
responsible for the overall electrical system design for high-voltage switchyards and power-plant
auxiliary systems. He is responsible for conceptual
plant designs, electrical equipment specifications,
and detailed engineering for large gas and steam
turbine power plants worldwide.
Mr. Schwartz is a Registered Professional Engineer in the States of
California and Colorado.

Paul S. Hamer (S78M79SM89) received the


B.S.E.E. degree from Virginia Polytechnic Institute
and State University, Blacksburg, in 1972 and the
M.S.E.E. degree from Oregon State University, Corvallis, in 1979.
He joined the Large Generator Department, Westinghouse Electric Corporation, in 1972 as a Service
Performance Engineer and was an Industrial Power
Systems Engineer and a Resident Engineer for the
Industry Services Division. Since 1979, he has been
with Chevron Corporation, Richmond, CA, where
he is currently a Senior Staff Electrical Engineer in the Projects and Engineering Technology Group of the Chevron Research and Technology Company.
His primary areas of responsibility include power systems, motors, and generator application and consultation. He is a member of the American Petroleum
Institute (API) Subcommittee on Electrical Equipment and represents the API
on the National Fire Protection Association Technical Committee on Lightning
Protection and the National Electrical Codes Code Making Panel 11.
Mr. Hamer is a Registered Electrical Engineer in the State of California.

Stephen R. Lambert (S65M69SM78F92)


received the B.S.E.E. and M.S. degrees from the
University of Illinois, Urbana, in 1969.
While at the University of Illinois, he taught
machine and basic circuit theory. In 1969, he joined
Commonwealth Associates, Inc., and was involved
in overvoltage studies, system planning, and training. He joined Power Technologies, Inc., Schenectady, NY, in 1973, where his prime responsibilities have been insulation coordination, substation
equipment appliations, transmission line design (insulation and field effects), analytical studies, and failure analysis. He has
performed numerous insulation coordination studies and has prepared many
functional substation equipment specifications with one project involving 3600
km of 765-kV line, 12 000 MVA of generation, and six 765-kV stations.
He developed statistical techniques to determine insulation probability of
flashover and has performed overvoltage studies on over 15 000 circuit km
of EHV transmission line. For 345-kV through 800-kV line designs, he
has specified conductor and bundle characteristics and configurations and
defined 60-Hz, lightning and switching surge insulation requirements, as well
as determined the lines electric field, audible noise, and RI response. He
has prepared and teaches courses and seminars on insulation coordination,
failure investigations, transient recovery voltage analysis, and transformer
concepts and applications. He is Vice Chairman of the PES Technical Council,
Past Chairman of the IEEE Switchgear Committee, the High-Voltage Circuit
Breaker Subcommittee, and the TRV WG. He has been appointed to the
IEEE Standards Board, is a member of the International Electrotechnical
Commission, WG 28.1, addressing Insulation Coordination, and the IEEE
WG on Insulation Coordination.
Mr. Lambert is a Registered Professional Engineer in the State of Michigan.

También podría gustarte