EHV SINGLE-POLE SWITCHING: IT IS NOT ONLY

A MATTER OF SECONDARY ARC EXTINCTION

B. Khodabakhchian
Abstract-- This paper reports the results of an in-dept
transient system study related to the application of single-pole
switching on the three sections of a 240 km double-circuit 330 kV
transmission line collecting some 1500 MW of Hydro power in
northern Qubec. While the paper discusses secondary arc
extinction time from laboratory test results, EMTP modeling of
secondary arc and related transient studies, the main focus is on
findings related to the creation of severe harmonic over-voltages
during both the dead-time and upon the circuit-breaker
reclosing. These findings were important enough to prevent the
application of single-pole switching on these lines. To the author's
knowledge, no such a restriction has ever been previously
mentioned in the literature on the application of single-poleswitching.
Keywords: Single-pole-switching - Secondary arc, Harmonic
over-voltages, EMTP-RV.
I. INTRODUCTION

YDRO-QUBEC bulk transmission system is composed

of eleven 735 kV very long lines (over 1000 km each)
bringing about 30,000 MW of hydro-electric power to its
customers in the south (Montreal and Qubec city). The
application of Single-Pole Switching (SPS) on these lines was
not judged to be justifiable firstly because of its marginal
benefit on the transient stability of a system with many
parallel lines and secondly because of the necessity to install
reactors at the neutral point of shunt inductances (costs). SPS
has however been applied since the mid 70's at the subtransmission voltage level of 330 kV where relatively long
double-circuit transmission lines (up to 160 km) connect the
main system to remote consumption sites containing some
small local generations. In these cases, the main role of SPS is
to maintain the transient stability of remote local generations
and to assure good continuity of service for remote customers.
This paper covers Hydro-Qubec's SPS experiences at 330
kV. It describes first the secondary arc extinction time
estimations obtained from laboratory tests and presents a
realistic secondary arc model in EMTP-RV. In the second
chapter, system studies related to the application of SPS on the
three sections of a 240 km double-circuit 330 kV transmission
line are described. Findings related to the latter are the main

Bahram Khodabakhchian is with System Studies Group, Hydro-Qubec

Transnergie, Complexe Desjardins, Tour de lEst, 10e tage, case postale
10000,
Montral
(Qc),
Canada,
H5B-1H7
(e-mail:
Khodabakhchian.Bahram@hydro.qc.ca).
Paper submitted to the International Conference on Power Systems
Transients (IPST2013) in Vancouver, Canada July 18-20, 2013.

purpose of this paper.

II. 330 KV SINGLE-POLE RECLOSING AT HYDRO-QUEBEC
Single-pole reclosing advantages and practice have been
described in the literature as early as 1942 in the US at 138 kV
[1] and later at 275 kV [2]. In early 70's, 500 kV field tests
were described for the first time in [3]. A good article on
various key factors determining the secondary arc extinction
times is presented in [4].
A. Laboratory tests
At Hydro-Qubec, single-pole reclosing times of 0.75 and
1 second have been applied systematically on 330 kV single
and double-circuit transmission lines respectively. In the early
80's, analysis of about 10 years of operation data showed some
unsuccessful reclosing due to the secondary arc non-extinction
on a 151 km compact 330 kV double-circuit transmission line.
As EMTP-based results of secondary arc parameters showed
in Table I indicated a higher than normal Vr/l parameter (17.4
kV/m vs. 13 kV/m), extensive laboratory tests were performed
at IREQ in 1986 [5]. During these outdoor tests, secondary arc
currents were varied from 10 to 40 A, recovery voltages from
7 to 18 kV/m, primary arc currents from 5 to 30 kA and line
compensation degrees from 0 to 50 and 75%.
TABLE I
SECONDARY ARC PARAMETERS OF HYDRO-QUBEC 330 kV
TRANSMISSION LINES (l = 2.3 m)
330 kV
line
Is
(A/km)
Vr (kV)
Vr/l
(kV/m)

Single circuit

Double-circuit (compact)

Transp.

Untransp.

Transp.

Untransp.

0.11

0.13

0.16

0.19

29.0

31.8

36.5

40

12.6

13.8

15.9

17.4

Is: Secondary arc current

Vr: Secondary arc recovery voltage
l: Insulator strings length
Figure 1 shows the test-deduced reclosing times on HydroQubec's 330 kV compact double-circuit lines. Depending on
the line length, the minimum short-circuit current and the
degree of compensation, single-pole reclosing times are
determined. For the 151 km transmission line in question,
reclosing time was elongated from 1 to 1.3 s and satisfactory
results were obtained.

Figure 3. Arc elongation for a 150 A secondary arc (from [7])

At this rate, EMTP-RV simulations of SPS on the

previously-mentioned 151 km 330 kV double-circuit line
predict a secondary arc extinction time of 0.91 s (Fig. 4) thus
confirming the adequate choice of a reclosing time equal to
1.3 s.

Figure 1. SPS reclosing times on 330 kV double-circuit compact lines

B. EMTP-RV Model
The 'Examples' library in EMTP-RV contains a secondary
arc model developed by the author which is based on the
pioneering work described in [6]. The user has only to input
the instant of fault creation, the initial arc length and its
variation with time. Figure 2 shows that an adequate arc length
variation with time is the key to reproduce precisely the field
recordings of secondary arc current and voltage. Such a fast
variation in the first 200 ms or so is typical and is mainly
dependant to the speed of the wind (around 10 km/h in the
test) and the amplitude of the primary arc current (around 5
kA rms). The so-called secondary arc 'instantaneous extinction
mode' [4] in this time frame is rarely seen in practice since a
high fault current and a low secondary arc current should coexist. Consequently, the model focuses mainly on the
'progressive extinction mode' [4] of the secondary arc which is
really of interest in practice. In this region, secondary arc
length variation of 5 pu/s has been reported (see Fig. 3).

Field Recording
(10-08-1986)

LAFORGE-2
289 MW

LAFORGE-1
817 MW

BRISAY
382 MW
SM

SM

SM

SM

Length Dependent arc model

EMTP-RV Simulation
(05-22-2005)

SM

SM

SM

The Brisay-Laforge2-Laforge1 hydro-electric complex in

northern Qubec generates about 1500 MW and was
developed in mid 1990's. This power is injected to the main
735 kV transmission system via a 240 km 330 kV doublecircuit transmission line (Fig. 5). Single-pole switching was
initially planned but never applied mainly because of a very
low lightning activity in the region (2-3 days per year) and the
fear of severe harmonic over-voltage creation.

SM

TRANSMISSION NETWORK

SM

III. SPS APPLICATION ON THE BRISAY - TILLY 330 KV

SM

Figure 4. Secondary arc extinction time at 330 kV predicted by EMTP-RV

1
2

Characteristic function

TILLY

5.5
2

4.5

735 kV
Network

3.5

2
1

3
735/315/15

CP

CP

CP

CP

107 kM

83 kM

YgYgD_np1

4
f(x)

55 kM

CP

CP

3
3

2.5

2
1.5
A

0.05

0.1

0.15

0.2

0.25

0.3
x

0.35

0.4

0.45

0.5

Figure 5. Brisay - Tilly 330 kV transmission network

0.55

Figure 2. EMTP-RV secondary arc model in comparison with a field test

Table II shows lightning-caused fault statistics on the 83

km line section between Laforge1 and Tilly. Between 1995

- Two generating units of 235 MVA at Brisay

- Four generating units of 147 MVA at Laforge1
- Two 55 km circuits between Brisay and Laforge2
- Two 107 km circuits between Laforge2 and Laforge1
- Two 83 km circuits between Laforge1 and Tilly.
A

linex_3172_3173

linedata1_3168_3169

linedata1_3166_3167

1
ZIN1
V_Nikamo
+ VM
?v

LINE DAT A
model in:

Z(f)

LINE DATA

linex_3172_3173_rv.pun

model in:

?z

V_LA2
+ VM
?v

V_Brisay
+ VM
?v

LINE DAT A

linedata1_3168_3169_rv.pun

model in:

linedata1_3166_3167_rv.pun

2 x 1650 MVA
755/315/12.5
YgYgD_np1
Slack: 745kVRMSLL/_0
Vsine_z:VwZ1

LF

Tilly

0.99/_0.7

2
1

LF4

BUS5
b
C1

10nF
VwZ1
755/315/12.5
YgYgD_np2

6.0e+ 5 /_-3
6.0e+ 5 /_-123
6.0e+ 5 /_117
Slack:LF4

R1
+
10k

SW2

107 km
GND
1.00/_6.3

FD

SW4
+
-1|1E15ms|0
b -1|173ms|0
-1|1E15ms|0
BUS10

SW14
C2

+
1E15ms|1E15|0
1070ms|1E15|0
1E15ms|1E15|0

SW5
+
-1|1E15|0

s30a
c

Brisay
55 km

1.01/_8.5
BUS14
a
c

FDline2
+

FD

b
b
c
a

b
b
c
a

SW6
+
-1|1E15|0

c
a

BUS15

SW7
+
-1|1E15|0

SW8
+
-1|1E15|0

BUS17

GND
c
b

FDline3
+

FD

BUS20

1.00/_13.1

c
b
a
a
b
c

a
a
SW10
b
+
c
-1|1E15|0
BUS18

SW12
+
-1|1E15|0

BUS21

LA1

SW13

10nF

a
c
c
a

LA2
83 km
FDline1

+
b
-1|1E15ms|0
-1|170ms|0
-1|1E15ms|0 BUS8

+
GND
100ms|300ms|0

1 circuit

SW15
+
1E15ms|1E15|0
1100ms|1E15|0
1E15ms|1E15|0

1 circuit

2 circuits

3
AVR_3

SM3
316.05/_6.4

RL_2
[R,L]

AVR_1
SM1

1
[R,L]

?m

-expci1
-pss2a
-hqrvw

SM

?m

GND

Vt

315/13.8

-30
N

RL3

expci1
partiel

LF

RL1
+

LF

-expci1
-pss2a
-hqrvn

SM

GND

-30

RL_1

AVR

YD_3
1
2

AVR

YD_1

GND
13.8kV
147MVA
LF1
PVbus:LF1

P=207MW
V=13.8kVRMSLL
SM:SM3

SPS was applied on the Tilly-Laforge1 circuit following a

single-line-to-ground fault at Tilly. Simulation results of
figure 7 indicate the creation of severe harmonic over-voltages
at the three substations during the SPS dead-time. Frequency
scan applied at Laforge2 shows indeed a near-third harmonic
resonance at this bus-bar when phase B is opened (Fig. 8).
These over-voltages are therefore created by the interaction of
the magnetizing currents of the power transformers at Tilly
(saturation following fault removal) with the high impedance
peak of third harmonic resonance. The fact that these overvoltages vanish gradually after several cycles is most probably
due to the de-tuning of the network third harmonic resonance
by the time domain variation of the machine impedances
(from X''d to X'd to Xd) following the fault removal.
Another configuration which was studied in detail is shown
in figure 9. This is the most frequent configuration which
corresponds to more than 1200 hours of operation during
2011. This scenario corresponds to:

expci1
partiel

GND

Phase:62

1 group of 235 MVA

1 group of 147 MVA

B

Figure 6. EMTP-RV design of the weakest configuration

Figure 6 corresponds to EMTP-RV simulation of the

weakest possible operation scenario for this network (129
hours of operation during 2011). This scenario corresponds to:
- One generating unit of 235 MVA at Brisay
- One generating unit of 147 MVA at Laforge1
- One 55 km circuit between Brisay and Laforge2
- Two 107 km circuits between Laforge2 and Laforge1
- One 83 km circuit between Laforge1 and Tilly.

Vt

LF3
13.8kV
235MVA
PVbus:LF3

P=120MW
V=315kVRMSLL
s7a
SM:SM1

TABLE II
FAULT STATISTICS ON THE 83 km DOUBLE-CIRCUIT LINE
BETWEEN TILLY AND LAFORGE1 SUBSTATIONS
Date
Faulted phases
Date
Faulted phases
(D-M-Y)
C#1 C#2
(D-M-Y)
C#1 C#2
30-06-95
a-- a-18-08-02
--c --30-06-95
a-- --01-08-05
a-c a-c
14-09-97
-b- --16-07-06
a-- -b21-06-99
--c --30-08-07
ab- ab22-08-99
--c --c
13-07-08
-b- --21-07-01
--c --c
10-08-08
a-- a-21-07-01
-b- -b03-07-11
-b- -b04-08-01
a-- ---

and 2011, a total of 15 fugitive faults have occurred on this

section. Out of these 15 cases, 9 faults have resulted in the
system separation and loss of generations of the whole three
generating stations. The amount of the total power loss varied
from 260 up to 900 MW. Out of these 9 cases, 7 could have
been prevented by the application of SPS. Consequently,
Hydro-Qubec Operation Department asked for a more
through SPS study with the aim to put the SPS in service, at
least in the most crucial section of Tilly-Laforge1. Precise
EMTP-RV simulations were therefore performed by using
frequency-dependant line and saturable transformer models.
Generation units were represented with synchronous machine
models along with their excitation and governor replica built
by using the control library elements of the Software. SPS
simulations were performed on many of the possible
configurations observed during the year 2011.

Figure 7. Harmonic over-voltages at 330 kV substations during SPS

dead-time

Figure 8. Frequency scan result at Laforge2 substation for the weak

configuration

In this scenario, a single-line-to-ground fault is applied on

phase A of both the circuits at Laforge1. Simulation results of
figure 10 show creation of fifth harmonic over-voltages at the
three substations (mostly on phase A), not during the SPS
dead-time but upon the circuit-breakers successful reclosing.
To understand the difference with the previous case, frequency
scans were again performed. Figure 11 shows that a fifthharmonic resonance exist normally at Brisay 330 kV

substation in this configuration. During the SPS dead-time,

this resonance is detuned due to the opened phase A on each
circuit. Upon circuit-breakers reclosing, saturation currents of
the machine transformers at Laforge1 and Brisay (following
fault removal) interact with the re-appearance of the firth
harmonic resonance, thus producing up to 1.8 pu of temporary
over-voltages for a couple of seconds. These over-voltages
were again judged to be harmful and consequently SPS was
prohibited permanently on the Brisay-Laforge2-Laforge1
network.

During the dead-time period, single-phase fault removal may

saturate power transformers at one of the two ends of the line.
The harmonic contents of these saturation currents may
interact with low order low-loss harmonic impedance (second
or third) of the network weakened by single-phase opening to
create severe temporary over-voltages. In stronger systems
where resonant conditions exist normally (mostly 4th, 5th and
7th), harmonic over-voltages may be created upon circuitbreakers reclosing.

To the author's knowledge, no such a restriction has been

ever reported in the literature. It is also the author's opinion
that the reported cases of harmful temporary over-voltage
generations are not specific to the Hydro-Qubec described
system. As SPS finds its most useful applications at EHV and
UHV (1100 kV in China for example), where one or two very
long transmission lines connect two systems through large
power transformers at each end, the same SPS restriction may
apply. Therefore, SPS application on such systems should
include extensive EMTP simulations both in time domain and
in the frequency scan mode.
A

linex_3172_3173

2 x 1650 MVA

Tilly

LF

Slack: 745kVRMSLL/_0
Vsine_z:VwZ1

3
10nF

0.99/_1.8

755/315/12.5
YgYgD_np2

+
R1
+
10k

FD

BUS8

3
10nF

C2

BUS14
SW7
a
+
c
-1| 0.17| 0
b
-1| 1E15| 0
-1| 1E15| 0
b
SW8
c
+
a
-1| 0.17| 0
-1| 1E15| 0
BUS17 -1| 1E15| 0

FDline2
+

FD

SW17
+
GND
95ms| 200ms| 0

SW13
+
1.2| 1e15| 0
1E15| 1E15| 0
1E15| 1E15| 0

2 circuits

SW16
+
1.2| 1e15| 0
1E15| 1E15| 0
1.01/_13.8
1E15| 1E15| 0

107 km

SW15
+

Figure 11. Frequency scan result at Brisay substation for the most frequent
configuration

V_Brisay
+VM
?v

LINE DATA
model in: linedata1_3166_3167_rv.pun

SW18
+
GND
95ms| 200ms| 0

a
BUS1
1.2| 1e15| 0
SW3 1.00/_9.4
b
1E15| 1E15| 0 c
+
+
1E15| 1E15|
0 b
a
-1| 1E15| 0
c
b
SW4 ?v
SW6
c
c
+
+
a -1| 1e15ms|
0
-1| 0.17| 0 a
b
-1| 1E15| 0
-1| 1E15|BUS10
0
BUS2

FDline1
+

V_LA2
+VM
?v

LINE DATA

6.0e+5 /_-3
6.0e+5 /_-123
6.0e+5 /_117
Slack:LF4

83 km

BUS5
b
a
c
c
SW2 ?v a
+
-1| 1e5ms| 0 b
SW1
+
-1| 1E15| 0

C1

VwZ1

Brisay

linedata1_3166_3167

model in: linedata1_3168_3169_rv.pun

-1| 0.17|
SW50
-1| 1E15| 0
-1| 1E15| 0

LF4

LA2

V_Nikamo
+VM
?v

LINE DATA
model in: linex_3172_3173_rv.pun

755/315/12.5
YgYgD_np1

linedata1_3168_3169

55 km

SW9
+
-1| 1E15| 0

BUS20
SW11
c
+
b
-1| 1E15| 0
a

FDline3
+

FD

1.00/_18.3

V.

SW12
+
-1| 1E15| 0
-1| 1E15| 0
-1| 1E15| 0
BUS9

a
SW10
b
+
-1| 1E15| 0 c
-1| 1E15| 0
BUS6
-1| 1E15| 0

a
b
c

SW14
+

2 circuits

1.2| 1E15| 0
1E15| 1E15| 0
1E15| 1E15| 0

2 circuits

3
AVR_3

SM3
RL_2

[R,L]

[R,L]

315.08/_9.6
+

m1
+VMYD_1
?v 1
2

AVR_1

SM1

AVR
?m

-expci1
-pss2a
-hqrvw

SM

-30

RL_1

Vt partiel
GND

P=415MW
V=315kVRMSLL
Phase:62
s36a
SM:SM3

The author would like to thank Hydro-Qubec Transnergie

for the permission to publish the results of this scientific work.
Appreciations go also to Yves Allard, Danielle McNabb and
Luc Grin-Lajoie for providing data and EMTP-RV support.

2 groupes de 235 MVA

4 groupes de 147 MVA

C

expci1
RL3
+

LF3
13.8kV
470MVA
PVbus:LF3

P=480MW
V=315kVRMSLL
s7a
SM:SM1

SM

LF

GND
13.8kV
588MVA
LF1
PVbus:LF1

LF

?m

-30

RL1
+

-expci1
-pss2a
-hqrvn

GND

Vt partiel

315/13.8

4 xfo

AVR

expci1

GND

LA1

YD_3
1
2

ACKNOWLEDGMENT

VI. REFERENCES

Figure 9. EMTP-RV design of the most frequent configuration

[1]
[2]
[3]

[4]
[5]
[6]

[7]

Figure 10. Harmonic over-voltages at 330 kV substations after SPS reclosing

IV. CONCLUSION
Beside secondary arc extinction, an important aspect of
SPS that needs to be studied in-dept is the potential of
harmonic over-voltage creation during both the dead-time
period and/or after successful circuit-breakers reclosing.

J.J. Trainor et al., "High-Speed Single-Pole Reclosing", AIEE

Transactions, Vol. 61, pp. 80-87, Feb. 1942.
K.H. Milne, "Single-Pole Reclosing Tests on Long 275 kV Transmission
Lines", AIEE Transactions, Vol. 61, pp. 658-661, Oct. 1963.
L. Edwards et al., "Single-Pole Switching on TVA's Paradise-Davidson
500 kV Line - Design Concepts and Staged Fault Test Results", IEEE
Transactions on PAS, Vol. 90, pp. 2436-2450, Nov. 1971.
H. -J. Haubrich et al., "Auto-renclenchement monophas dans les
rseaux trs haute tension ", CIGRE Report31-09, Paris, 1974
R.J. Rajotte, "tude de l'arc secondaire pour le renclenchement
monophas 330 kV", IREQ Internal Report 4117G, Jan. 1988.
M. Kizilcay et al., "Interaction of the Secondary Arc with the
Transmission System during Single-Phase Autoreclosure", PowerTech
Conference Proceedings,IEEE Bologna, 23-26 June 2003.
A. Cmara et al., "Single-Phase Auto-reclosure Studies: Secondary Arc
Model Research Including a 500 kV Line Experimental Circuit",
Proceedings of the 2008 International Conference on HV engineering
and Applications, Chongqing, China, Nov. 9-13, 2008.