IEEE Transactions on Power Delivery, Vol. 11, No.

2,April 1996 865

Mitigation of Circuit Breaker Transient Recovery

Voltages Associated with Current Limiting Reactors

D.F. Peelo P. Diamanti R. Beauchemin

G.S. Polovick R. Presta
J.H. Sawada A. Sarshar

B.C. Hydro Trench Electric H ydro-Quebec

Abstract - B.C. Hydro and Hydro-Quebec utilize current APPLICATION OF SERIES REACTORS
limiting reactors on medium voltage distribution feeders to limit
the magnitude of fault current. The reactors make a significant The application of series reactors in substations is described
contribution to the severity of the transient recovery voltage in detail in the reference, but a brief description is provided
experienced by feeder and bus circuit breakers on clearing for background information.
feeder faults. Based on the conclusions of an investigation of
actual circuit breaker failures while performing this duty, the
mitigation of the transient recovery voltage associated with the The purpose of series reactors is to limit the magnitude of
reactors is described. fault current. The reactors can be applied in a number of
locations in order to achieve this purpose, on main buses, on
INTRODUCTION feeders or in transformer neutrals. Both B.C. Hydro and
Hydro-Quebec apply series reactors on individual feeders
B.C. Hydro and Hydro-Quebec medium voltage substation (Fig. 1).
arrangements are characterized by the use of air-cored current
limiting series reactors on feeders. Both utilities experienced Large capacity substations are commonly used to supply
a number of SF, puffer type circuit breaker failures. distribution systems. The associated low voltage bus short
Investigation showed that the failure of these particular circuit circuit levels can be very high. If not controlled, these high
breakers could be attributed in part to the transient recovery short circuit levels are imposed on all distribution feeders with
voltage (TRV) component associated with the series reactors. undesireable consequences including:

The application of series reactors on medium voltage systems 1. feeder circuit breakers and associated buswork and
is discussed briefly for background information. The paper disconnect switches must be rated for the bus short-
then describes the nature of the failures and the subsequent circuit level
failure investigation, which included actual TRV
measurements on both utility systems and EMTP studies using 2. preventing the use of some downstream devices such
models based on the measurements, and its conclusions. as reclosers and fuses because the short circuit duty
may be excessive
The investigation showed that TRV control measures needed
to be taken. Details of these control measures, including the 3. transformers will be exposed to higher fault levels and
development of a dedicated capacitive device to apply across a high frequency of such faults (given that the feeder
the series reactor, are described. fault is the most common fault on power systems),
resulting in possible reduction of transformer life
expectancy
95 SM 498-6 PWRD A paper recommended and approved
by the IEEE Switchgear Committee of the IEEE Power
Engineering Society for presentation at the 1995
4. possible violent failure of distribution transformers
IEEE/PES Summer Meeting, July 23-27, 1995, Portland, unless current limiting fuses are applied on each
OR. Manuscript submitted December 24, 1994; made transformer
available for printing April 27, 1995.
5. burndown of aerial conductors

Q 1995 IEEE
0885-8977/96/$05.00
866

P 25YTT1

Fig. 1 Typical B.C. Hydro 25 kV One-Line

6. joint use of poles for power distribution and to 12.5 kA when two feeders are temporarily paralleled to
communication circuits may be precluded. transfer load from one feeder to another.

The use of series reactors on feeders mitigates most or all Fewer bus reactors could be applied, but such reactors are
of these effects. Important considerations when using series not attractive from a regulation point of view (as are high
reactors include cost, space requirements and installation cost, impedance transformers) and watts loss. Transformer neutral
watt and var losses and regulation through the reactor (not reactors are also unattractive, being effective only for SLG
significant if the load power factor is kept above 0.9). The faults.
reduction in fault level is, however, significant with clear
advantage. At 25 kV, B.C. Hydro applies 400 A, 5 mH Both utilities use 12 kA circuit breakers on feeders, while
reactors. This results in a feeder SLG fault level of maximum bus circuit breakers are rated at 20, 25, 31.5 or 50 U.
6 kA,compared to a level in excess of 20 kA if the reactors
were not applied. Hydro-Quebec similarly apply 25 kV series CIRCUIT BKEAKER FAILURES
reactors in the range 2.1 to 3.5 mH dependent on transformer
size and impedance with the intent of limiting the fault level The Hydro-Quebec circuit breaker failures were associated
with faults on cable circuits, while those at B.C. Hydro were
 867

associated with overhead distribution feeder faults. Only the Current magnitude: The fault current levels were typically
latter faults are discussed. in the range around 2 to 4 kA and thus were a very low
percentage fault compared to the 25 kA rating of the circuit
The B.C. Hydro circuit breaker failllres occurred over a nine breaker (a 12 kA rated circuit breaker was specified, but a
month period. The failures were the result of five faults and 25 kA rated circuit breaker was supplied). This issue was
are detailed below: initially ruled out as a factor, but was later concluded to be a
contributing factor.
-
Cause Fault Type
TRV: This was considered the most probable cause,
Falling tree Phase-to-phase however, it remained to be explained why this was the case.
Motor vehicle accident Three-phase As an interim solution, surge capacitors were installed
Falling tree Phase-to-phase between the series reactors and the circuit breakers, on the
Unknown unknown main buses and between the transformer and the bus circuit
Falling tree Single-phase to ground breakers.

Compared to laboratory type test conditions, the conditions INVESTlGATION AND SOLUTION
in the field appeared to be less onerous. For example, circuit
breaker gas pressure in the field was rated gas pressure versus An intensive investigation determined that one contributing
lockout gas pressure for the laboratory tests; the maximum factor to the circuit breaker failures was indeed the TRV. The
fault level in the field was 4 kA at 60 Hz versus 6.9 kA at difference between the field TRV and the laboratory TRV was
50 Hz for the laboratory tests; and only one failure was in the time delay. In the field, the time delay was
associated with the laboratory three-phase ungrounded test demonstrated to be about 0 . 7 , or ~ less, most probably the
condition. Note that B.C. Hydro was aware of the TRV latter, while the time delay associated with laboratory tests
imposed by the series reactors, having done a significant was about 2 ps. Other evidence clearly demonstrated that the
number of such measurements (a typical measurement is fault level relative to the circuit breaker short circuit rating
shown in Fig. 2, while a similar Hydro-Quebec measurement was the second contributing factor to the failures.
is shown in Fig. 3). The specified TRV was a four-paramater
TRV to reflect the initial part due to the series reactor and the The solution was to control the TRV. The TRV, as noted
later part due to the overall bus and transformer arrangement. earlier, consists of an initial fast transient due to the series
The type tests on this basis were successful and yet the field reactor and a later slower part due to the overall bus
failures occurred. arrangement and the transformer or transformers (refer to
Figs. 2 and 3). Control of the TRV, therefore, had to take
A number of possible causes were considered. Briefly, care of the fast transient for both the feeder and bus circuit
these were: breakers (a bus circuit breaker clearing a feeder fault also sees
the fast transient); and it also made sense to consider control
Evolving fault: This cause was considered a possibility at of the slower part of the TRV in order to be able to specify
the time of the first failure but was abandoned once the a single TRV requirement for all circuit breakers. B.C. Hydro
second failure occurred. opted for total control of the ‘TRV as described below.
Hydro-Quebec chose to control only the fast part of the TRV
Layout arrangement: The issue here was whether the on the basis that standard TRV values in accordance with
close proximity of the series reactor and its associated JEC 56 covered the slower part of the TRV on their system
magnetic field were influencing the ability of the circuit with respect to the bus circuit breakers.
breaker to interrupt the current. Review eliminated this
possibility as a failure cause. The traditional method of controlling TRVs is to install
surge capacitors. In this case, there are a number of options
Grounded versus ungrounded faults: Field faults tend to as to location of the capacitors, each with advantages and
involve ground (varying from low to high resistance), whereas disadvantages.
laboratory tests are performed on a three-phase ungrounded
basis. From the point of view of the circuit breaker, this is The first option is to apply surge capacitors to ground. This
reflected in lower TRV values, but slightly longer arcing option has the advantage that it eliminates the fast part of the
times. However, laboratory tests with forced prolonged arcing TRV, reduces the overall frequency of TRV and hence the
times eliminated this possibility. RRRV, and increases the time delay significantly; the
disadvantages are that the peak value of the TRV is increased,
the capacitors would be required at three locations (between
868
EMTP studies were performed to derive the required surge
capacitor values. Based on the actual field TRV
measurements, both utilities used validated models for this
work. An example of the output from a Hydro-Quebec study
is shown in Fig. 4, while a composite example from a
B.C. Hydro study is shown in Fig. 5 (compare to Figs. 2 and
3). For the surge capacitor to ground, a commercially
available unit of 0.125 pF was selected. For the capacitive
device across the series reactor, a value of 30 nF was derived.
This vaiue was subsequently reduced to 20 nF for physical
sizing reasons without having any significant effect on TRV
control performance.

40 -

35 -

0 10 20 30 40 50 60 30 -

Time (ps) -
225 -
a
U)

r
> -
Fig. 2 B.C. Hydro TRV Measurement (first pole 0 20

three-phase ungrounded fault) on 25 kV Feeder 15 -

the series reactors and the feeder circuit breakers, on the main
buses and on the transformer secondaries - all highly
undesireable from a station layout point of view) and the
associated charging currents have to be interrupted by the bus
1:3 1.$5 1:4 1.15 1:5 1.k
and feeder disconnect switches. The second option is to apply Time (ms)
the surge capacitor across the circuit breaker. The advantage
here is that it does increase the time delay and may have
some impact on the overall TRV dependent on size; the
disadvantages are that it would be difficult for the utility to Fig. 3 Hydro-Quebec TRV Measurement (first pole
specify, it may need to have quite a high capacitance value in three-phase ungrounded fault) on 25 kV Feeder
order to provide oveiall TRV control and furthermore creates
the possibility of series resonance with customer inductive CAPACITIVE DEVICE
loads. The third and last option is to apply the surge
capacitor across the series reactor. The advantages of this Design and test requirements for the reactor capacitive
option are that it eliminates the first part of the TRV, is device were determined with a view to ensuring that the
stressed only with high voltage during faults and could withstand capability and operational integrity of the series
possibly be located inside the reactors and thus have no reactor assembly was not impaired. The device was,
impact on station layout. therefore, designed and tested to successfully operate in this
most stringent application.
No one of the surge capacitor installation options alone met
the degree of TRV control desired by B.C. Hydro. To Design requirements for the capacitive device included the
achieve the desired degree of TRV control, a combination of following:
the first and third options is necessary, namely, apply surge
capacitors to ground on the transformers and across the series 1. The device was to be installed inside the reactor
reactors. The next step was to size and specify the required without having any adverse effect on the reactor.
capacitor units, which included the development of a
capacitive suitable for installation inside the reactors.
 869

Fig. 4 TRV across circuit breaker calculated for three phases to ground fault fed
by one, two and three 140 MVA transformers connected in parallel

2. The device had to be suitable for outdoor and indoor testing, some failures occurred. Design modifications were
applications with UV and track-resistant housing made to address these failures and the final design
material over an operating temperature range of -50°C successfully passed the type test program.
to 125°C.
The type test program was developed jointly by B.C. Hydro,
3. The device had to be capable of operating in the Hydro-Quebec and the reactor manufacturer. The program
magnetic field inside the reactor. included the following tests:

4. The main electrical characteristics of the device were Temperature rise test: This test verifies the operation of
specified as: the capacitive device in the magnetic field and higher than
ambient temperature when installed inside the reactor.
Maximum rated voltage 16 kV
Nominal rated voltage 1.5 kV Rated short-time fault current test: This test verifies the
Impulse withstand voltage 125 kV capability of the device and its mounting arrangement to
1.2150 ps and 10-20/2500ps withstand the forces due to short-circuit current through the
One minute 60 Hz withstand level 50 kV reactor.
Capacitance 20 nF
Special thermal stability test: This test, applicable to
Various dielectric materials were considered and evaluated indoor installations, verifies the capability of the device to
during the development of the device. withstand the highest operating range temperature.

It was found that, due to the stringent operating conditions Withstand tests: These tests, applicable to outdoor
only high temperature, high quality dielectric materials would installations, verifies the capability of the device to perform
perform successfully in the application. Initial testing at room under the most severe service conditions. The tests included
temperature were favourable, but when the device was an overvoltage test, which incorporated temperature cycling to
subjected to extremes in temperature and extensive dielectric -50°C and a total of 450 voltage withstand tests at 16 kV.
870

voltage withstand tests -- starting with a capacitive and

Y BUS CE TRVs: WITH (L WITHOUT TRV CAPACITORS dissipation factor measurements at 16 kV and finishing with
those same measurements followed by a partial discharge
measurement.

CONCLUSIONS

The application of current limiting series reactors on

distribution feeder circuits has significant technical and
economic merit. The TRV imposed by the reactors can be
effectively eliminated by means of a capacitor applied across
Q I S x 1 U 6 0 7 s 9 4
the reactors. Both B.C. Hydro and Hydro-Quebec have
TI. ("sac) adopted this solution, thus allowing the use of standard circuit
breakers with some additional short-circuit testing in addition
to that required by the standards.

REFERENCES

K.K. Nishikamara, "Application of Current Limiting

Reactors in Substations". Presented at the Canadian
Electrical Association Spring Meeting, March 1980 in
Montreal.

M. Landry, "Mesures de la tension transitoire de

rdtablssement sur le rCseau de distribution
d'Hydro-QuCbec",Rapport IREQ 4RT39896, Hydro-
" I
1 5 x 1 4 5 6 0 7 5 9 0
t Quebec, June 1987.
Q 105 Ifl
Tlm turecl

BIOGRAPHIES

FEEDER CB TRVs: WITH & WITHOUT BUS & RX TRV CAPACITORS D.F. Peelo received his B.E. degree from University College
g ..............................................................................
I I s
Dublin in 1965. He joined ASEA after graduation working in
the High Voltage Laboratory of the Power Transmission
Products Division in Ludvika, Sweden. Jn 1973 he joined
B.C. Hydro working first in the Quality Control and
Inspection Department and later in the Stations Engineering
Division, where he is currently a Specialist Engineer
responsible for the procurement and application of switchgear
and surge arresters. He is a Fellow of the TEE and a Senior
Member of the IEEE, and at present is the Vice-Chair,
Standards Coordination of the IEEE PES Switchgear
Til. luus)
Committee.

G.S. Folovick received his B.E. degree from the University

Fig. 5 Bus and feeder circuit breaker TRVs €or three-phase of Saskatchewan in 1910. He joined Westinghouse Canada as
ungrounded 25 kV feeder fault fed b y two transformers. a power transformer design engineer. In 1974 he joined
B.C. Hydro working first in the Quality Control and
Dielectric tests: These tests verified the dielectric withstand Inspection Department and later in the Stations Engineering
capability of the device for both 60 Hz and impulse test Division, where he is currently a Specialist Engineer
voltages under dry and wet conditions as applicable. responsible for the procurement and application of reactive
equipment. He is a Member of the IEEE and has worked on
A routine test program was also agreed upon. The program numerous CEA and CSA working groups related to reactive
consists of a sequence of tests -- including 60 Hz and impulse equipment standards.
 87 1

J.H. Sawada joined B.C. Hydro in 1976. For the last has written and presented various papers on the subject of line
14 years, he has been with the Analytical Studies group in the traps and insulation coordination criteria. He is an active
System Planning Division. He received his BASc(72) and member of NEMA Working Group ASC C93 Power Line
MASc(74) both in Electrical Engineering from the University Carrier Committee and a member of IEEE.
of British Columbia. His main interest is in the insulation
coordination field. His current role is to provide technical R Presta graduated in Electrical Engineering, Power Option
services in the power system transients analysis area and from Ryerson Polytechnical Institute, Toronto, Canada in
EMTP applications. 1979. In the same year, he joined Trench Electric as a
Design Engineer specializing in high impulse coils. In 1988,
A. Sarshar was born in Iran in 1963. He received his B.Sc. he became the Test Lab Manager.
in Electrical Engineering from Purdue University in 1987 and
his Masters of Electrical Engineering from the University of R. Beauchemin was bom in Quebec, Canada. He received
Toronto in 1991. He worked for Federal Pioneer Ltd. from his B.Sc.A degree in Electrical Engineering in 1972 from
1990-1992 as Design Engineer. Since 1992 he has been Ecole Polytechnique de Montreal. He has been working as
employed in the R&D group of Trench Electric as an Engineer for Hydro-Quebec since 1972 in substation design
Application Engineer. His study experience has been focused and equipment standardization. While with the Transmission
on distribution and transmission system modeling using the Equipment Department, he was responsible for the
Electro Magnetic Transients Program (EMTP). development of the Hydro-Qu6bec standards on HV and EHV
circuit-breakers and their certification. In 1990, he joined the
P. Diamanti is the Engineering Manager in charge of the Power Network Planning Department. He is involved in
Line Traps Engineering and Drafting at Haefely-Trench. He studies to fix electrical requirements of HV apparatus. He is
graduated in 1975 in Electrical Engineering, Power Option, a registered Professional Engineer in the province of QuCbec
from Ryerson Polytechnical Institute, Toronto, Canada. He and a member of the CSA Power Capacitors Committee.