Electrical Power and Energy Systems 45 (2013) 340–345

Contents lists available at SciVerse ScienceDirect

Electrical Power and Energy Systems

journal homepage: www.elsevier.com/locate/ijepes

Side effects of Current-Limiting Reactors on power system

Shuiming Chen ⇑, Pengcheng Yang, Wei Xu, Xiao Bian, Wei Wang
State Key Lab of Control and Simulation of Power Systems and Generation Equipments, Tsinghua University, 100084 Beijing, China

a r t i c l e i n f o a b s t r a c t

Article history: Air-cored Current-Limiting Reactor (CLR) has already been widely applied to limit the magnitude of fault
Received 16 December 2010 current in power systems in many countries, Brazil for example, and meanwhile the CLR has made a sig-
Accepted 11 July 2012 nificant contribution to the severity of transient overvoltages in power systems, such as transient recov-
Available online 26 October 2012
ery voltage (TRV), switching overvoltage (SOV), and temporary overvoltage (TOV). The present paper here
did a lot of simulations on the application of CLR for a real-life transmission line, suggesting that with
Keywords: equipment parameters of originally draft, TRV would have exceeded the circuit breakers’ TRV ratings
Current-Limiting Reactor
and necessary measures should be taken to decrease TRV. Digital simulations were made with PSCAD/
Transient recovery voltage
Switching overvoltage
EMTDC software and the results implied that the magnitude of the SOV and the TOV increased up to
Temporary overvoltage or even exceeded the safe limits with the CLR installed. Relevant studies are strongly recommended
before the CLR is applied to ensure the safety of equipment and transmission system.
Ó 2012 Published by Elsevier Ltd.

1. Introduction transient recovery voltage (TRV) across the circuit breaker con-
tacts, associated with the interruption of the circuit current [3,4].
With the development of power system, the short-circuit cur- The switching overvoltage (SOV) and the temporary overvoltage
rent levels are increasing rapidly. In some substations, the fault (TOV) will increase as well because of the lumped inductance. In
current has been close to or even exceeded the capacity of the cir- this presented research, various simulations on those overvoltages
cuit breakers. Decreasing the short-circuit current levels can not were carried out with PSCAD/EMTDC, and the results are discussed
only prevent important equipment from being broken down by in the following sections.
large fault current, but also reduce electromagnetic interference
caused by fault current. The most common solutions to high fault
2. Side effects of the CLR on the TRV
current levels are [1]: up-rating of switchgear and other equip-
ment; splitting the grid and introducing higher voltage connec-
During the short-circuit current interruption process, at current
tions (AC or DC); introducing higher impedance transformers and
zero when current is interrupted, the system oscillates in accor-
series reactors; and using complex strategies like sequential net-
dance with its natural frequency. This system-generated voltage,
work tripping. Nevertheless, these alternatives may bring other
called the TRV, is impressed across the opening breaker contacts
problems such as loss of power system security and reliability,
and stresses the gap insulation. If the TRV has a high rate of rise
high costs, and increase of power losses. The applications of air-
in the first few microseconds following current interruption, it
cored Current-Limiting Reactor (CLR) in Brazil [1] and Canada [2]
may reestablish a current flow sufficient to heat the arc column
implied that CLR had significant effects on decreasing fault current,
and restore conduction. If the heating exceeds the breakers capa-
thus, up to now, the usage of CLR is a widely adopted solution in
bility to remove heat and cool the arc channel, then reignition
the world.
occurs.
The guidance factors for the CLR dimensioning, which could
When the CLR is inserted into the circuit, for instance, to limit
even make the CLR application unfeasible were the voltage drop,
the short-circuit current, the rise rate of TRV tends to drastically
the joule losses and the high magnetic fluxes (higher distances/
increase because of its very large surge impedance. Fortunately,
clearances required). On the other hand, despite these disadvan-
installing a suitable capacitor across the reactor may easily solve
tages, their effects could be economically compensated when
this problem [5–8]. Capacitors are necessary in some cases and
avoiding equipment substitution.
usually an in-depth computer analysis must be performed to make
It is demonstrated that the presence of a lumped inductance in
sure the circuit breaker characteristics are not exceeded [9–11].
an electric circuit will lead to an increase in the severity of the
In China, the CLR is installed for the first time on 500 kV
Sijing–Huangdu transmission line shown in Fig. 1 and Table 1
⇑ Corresponding author. Tel.: +86 10 62772197; fax: +86 10 62784709. [12]. Intensive investigations show that the TRV of this line
E-mail address: [email protected] (S. Chen). may lead to the circuit breaker failures. The specified TRV was a

0142-0615/$ - see front matter Ó 2012 Published by Elsevier Ltd.

http://dx.doi.org/10.1016/j.ijepes.2012.07.062
 S. Chen et al. / Electrical Power and Energy Systems 45 (2013) 340–345 341

Shipai
Huangdu a

Sijing

Nanqiao
Huxi

Fig. 1. Equivalent network.

b
Table 1
Transmission line parameters of equivalence power network.

Line names Wires Length (km)

Shipai–Huangdu Double-circuit 4  400 38
Huangdu–Sijing Double-circuit 4  400 23
Sijing–Nanqiao Double-circuit 4  400 26
Huxi–Sijing Double-circuit 6  630 25

four-parameter TRV to reflect the initial part due to the series reac-
tor and the later part due to the overall bus and transformer
arrangement, the parameters of TRV are listed in Table 2 [13,14]. Fig. 2. (a) TRV at HD. (b) TRV at SJ (dash red line: the limit of TRV, dot green line:
single phase fault at HD, dash dot blue line: single phase fault in the middle, solid
2.1. TRV caused by single phase short-circuit black line: single phase fault at SJ). (For interpretation of the references to colour in
this figure legend, the reader is referred to the web version of this article.)

The magnitude and shape of the TRV vary with the type of
grounding and the point where the fault takes place. Single phase
short-circuit is the most frequently encountered problem in the
power system, thus TRV caused by it is of great importance to
a
the safety of circuit breakers. The calculated TRV at different loca-
tions of the transmission line are shown in Fig. 2a for Huangdu
(HD) substation and in Fig. 2b for Sijing (SJ) substation.
Fig. 2 illustrates the oscillograph of voltage versus time during
an interruption. Notice the TRV at both substations exceed the lim-
it while single phase fault takes place. Especially at Sijing substa-
tion, the steepness of TRV is much larger than the limit.

2.2. TRV caused by three phases short-circuit

This section examined the three-phase grounded fault which

was determined by analysis to impose a severer TRV than the b
single-phase-to-ground fault in this case. Note that the three-
phase ungrounded fault was considered too rare of an event for
the 500 kV system under study. Extensive related researches have
been done and the TRV at both sides of transmission line with dif-
ferent fault locations are shown in Fig. 3.
As shown in Fig. 3, the severity of the TRV caused by three
phases fault increases in both the magnitude and the steepness.
The TRV at both substations exceed the limit greatly and necessary
measures should be taken to reduce the TRV.

2.3. Measures to decrease TRV

Fig. 3. (a) TRV at HD. (b) TRV at SJ (dash red line: the limit of TRV, dot green line:
three phases fault at HD, dash dot blue line: three phases fault in the middle, solid
Based on prior concerns for the TRV, the resolution to cases black line: three phases fault at SJ). (For interpretation of the references to colour in
where the TRV exceeds the circuit breaker rating may be the this figure legend, the reader is referred to the web version of this article.)

Table 2
Standard value of TRV with four-parameters in 500 kV.

Rated voltage (kV) Test method First off coefficient Amplitude factor First reference voltage (kV) Time1 (ls) Amplitude of TRV (kV) Time2 (ls)
550 Outlet terminal fault 1.3 1.4 438 219 817 876
Short line fault 1 1.4 337 168 629 672
342 S. Chen et al. / Electrical Power and Energy Systems 45 (2013) 340–345

simple addition of surge capacitors. In the original project draft, tions and further analysis was performed to determine the
60 nF capacitors were planned to be installed at SJ substation to re- minimum capacitance required for the circuit breakers to mitigate
duce the TRV. The effects of the CLR and the surge capacitors on the TRV resulting from faults. In this case, the method that 1 lF
TRV are studied and the results in different conditions are shown and 0.8 lF parallel capacitors installed at HD and SJ substation
in Fig. 4. respectively works well and the TRV at both substations meet
To summarize, Fig. 4 demonstrates that the CLR and capacitors the limit as shown in Fig. 5.
have great effect on TRV, and the following conclusions can be
drawn:
3. Side effects of the CLR on switching overvoltages

(1) Without the CLR and the surge capacitors, TRV at HD are
3.1. Theoretical analysis
lower than the limit, while the steepness of TRV at SJ
exceeds the limit.
As the insulation level of EHVAC systems is largely determined
(2) The imposed high-frequency TRV caused by CLR installed at
by the magnitude of switching overvoltages, it is necessary to
SJ exceeds the circuit breaker TRV capability.
recalculate the voltage stresses on aging equipment still in service
(3) Capacitors installed at SJ in the original draft have some
before the CLRs are installed. Theoretical analysis had indicated
effects on decreasing the TRV at SJ. Both the amplitude and
that switching overvoltages will increase with CLR applied. For a
steepness of TRV reduce. While at HD, there is little effect
three phase transmission line, the three phases energizing switch-
and the TRV scarcely change.
ing overvoltage is [15]:
(4) With CLR and the surge capacitors in the original draft, TRV
at both substations exceeds the limit. Additional methods to uc ðtÞ ¼ U m ½ cosðxt þ uÞ
reduce TRV should be taken. qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
þ ðx=x0 Þ2 sin u þ cos2 u  edt cosðx0 t þ hÞ ð1Þ
For this study the focus is on controlling the component of TRV
where u = u0  arctan[(xL  1/xC)/R], tan h = x/x0  tan u, x0 ¼
to meet specified short-line fault TRV capabilities for circuit break- qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ers. Larger capacitors are suggested to be installed at both substa- 1=LC  d2 , d = R/2L, uc is the voltage at the terminal of the trans-

a d

b e

c f

Fig. 4. Single phase fault with 60 nF capacitors installed at SJ. (a) TRV at HD while fault near HD. (b) TRV at SJ while fault near HD. (c) TRV at HD while fault in the middle. (d)
TRV at HD while fault in the middle. (e) TRV at HD while fault near SJ. (f) TRV at SJ while fault near SJ (dash red line: the limit of TRV, dash dot blue line: no THE CLR and
capacitor; dot green line: with THE CLR; solid black line: with THE CLR and capacitor). (For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)
 S. Chen et al. / Electrical Power and Energy Systems 45 (2013) 340–345 343

made to get the 98% probability overvoltage. The results are listed
a
in Tables 4 and 5.
These results show that adding CLR to the transmission line in-
deed increases the switching overvoltages, which is consistent
with the theory above. Since the impedance of the CLR used in
the simulations is very small compared to the impedance of the
transmission line, the rising rates of switching overvoltages are
not obviously. Three phases energizing and single phase reclosing
switching overvoltages rise by 2.01% and 1.53% respectively. The
severest switching overvoltage is 1.79 p.u., just a little lower than
the limit value (1.8 p.u.) in China.
Based on Eqs. (2) and (3), larger impedance CLR could cause
greater overvoltages on the transmission line. CLRs with other
b impedances were also used in the simulations and the results are
shown in Table 6.
As stated in Table 6, it is demonstrated that installing CLR will
lead to an increment of the switching overvoltages. The up-rates
of overvoltages depend on the impedance of CLR, especially for
the three phases energizing switching overvoltage. If the imped-
ance of CLR is larger than 20 X, the overvoltage will exceed the
limit (1.8 p.u.).

3.3. Equivalent Circuit Method

Switching transients usually exhibit complex waveforms for

Fig. 5. Single phase fault while measures are taken (a) TRV at HD and (b) TRV at SJ
(dash red line: the limit of TRV, dot green line: single phase fault at HD, dash dot which the fundamental frequency usually lies in the range
blue line: single phase fault in the middle, solid black line: single phase fault at SJ). 100–1000 Hz and in some cases a very steep voltage rise or col-
(For interpretation of the references to colour in this figure legend, the reader is lapse can occur, thus it is difficult to establish a accurate model
referred to the web version of this article.)
of the CLR in PSCAD/EMTDC. A much simpler method called Equiv-
alent Circuit Method was developed to estimate the increment of
mission line, u0 is the switching angle, R, L and C are equivalent switching overvoltage caused by CLR. As shown in Eqs. (2) and
resistance, inductance and capacitor of transmission line respec- (3), the increase rate depends on L. The CLR leads to the total
tively. impedance of line rises as if the transmission line is prolonged.
For the sake of simplicity, R can be neglected because it is very Correspondingly, the CLR can be modeled as a segment of trans-
small compared to L. As a result, the attenuation coefficient d mission line in the simulations for the sake of simplicity.
equals to 0. Assuming that u0 is 90° (the severest situation) and
xL  1/xC, the following formulas can be derived: Table 4
Three phase energizing switching overvoltages.
U cm ¼ U m =ð1  x2 =x20 Þ ð2Þ 0 km 80 km 160 km 240 km 320 km 400 km

pffiffiffiffiffiffi No CLR (p.u.) 1.49 1.61 1.68 1.72 1.76 1.75

x0 ¼ 1= LC ð3Þ With CLR (p.u.) 1.52 1.64 1.70 1.74 1.79 1.75
Rising rate (%) 2.01 1.86 1.19 1.16 1.70 0
As shown in Eqs. (2) and (3), larger L could lead to greater Ucm. If
the CLRs are installed on the transmission lines, the equivalent
inductance L will rise up and undesirable overvoltages in transmis-
sion lines may be caused. Related simulations are carried out and Table 5
Single phase reclosing switching overvoltages.
the results are shown in the following section.
0 km 80 km 160 km 240 km 320 km 400 km
3.2. Simulation results No CLR (p.u.) 1.25 1.31 1.34 1.34 1.32 1.29
With CLR (p.u.) 1.25 1.33 1.36 1.36 1.34 1.29
Rising rate (%) 0 1.53 1.49 1.49 1.52 0
In this section, the study was performed using a real 750 kV
power transmission line with the configurations shown pffiffiffi pinffiffiffiTable 3.
The normal peak value of any phase voltage is 800 2= 3 kV and
this value is taken as base for voltage p.u. The digital simulations Table 6
were made using PSCAD/EMTDC software, as it have been proved Switching overvoltage with different THE CLR.

to be the highest and most accurate digital simulation environment CLR (X) 0 km 80 km 160 km 240 km 320 km 400 km
for handling such a problem. Three phase energizing
Since the amplitude of switching overvoltages depends on the 7.85 (p.u.) 1.52 1.64 1.70 1.74 1.79 1.75
switching angle, Monte Carlo Method based simulations were 20 (p.u.) 1.55 1.66 1.73 1.76 1.82 1.76
40 (p.u.) 1.60 1.70 1.76 1.81 1.86 1.76
60 (p.u.) 1.64 1.76 1.85 1.89 1.90 1.77
Table 3 Single phase reclosing
Parameters of 750 kV transmission line. 7.85 (p.u.) 1.24 1.33 1.36 1.36 1.34 1.29
20 (p.u.) 1.23 1.37 1.38 1.40 1.39 1.28
Line names Wires Length Load
40 (p.u.) 1.23 1.42 1.42 1.47 1.47 1.27
Yinchuan–Lanzhou 6  400 400 km 413 MW + 571 MVAR 60 (p.u.) 1.24 1.44 1.44 1.49 1.48 1.29
344 S. Chen et al. / Electrical Power and Energy Systems 45 (2013) 340–345

Table 7
Switching overvoltages with different calculation methods.

Methods 0 km 80 km 160 km 240 km 320 km 400 km

Three phases energizing
Line + CLR (kV) 992 1069 1113 1138 1167 1145
Equivalent Circuit (kV) 954 1056 1114 1145 1159 1146
Single phase reclosing
Line + CLR (kV) 809 871 885 890 877 839
Equivalent Circuit (kV) 826 890 903 892 902 860

Assuming that x is the oscillation frequency of line with CLR, Table 8

and x0 is the oscillation frequency of the equivalent line. In the Temporary overvoltages.
condition that x = x0 , the following equation can be derived:
Methods 0 km 80 km 160 km 240 km 320 km 400 km

ðL0  l þ LC Þ  C 0  l ¼ ðl þ le Þ2  L0 C 0 ð4Þ Ferranti effect

No CLR (p.u.) 1.10 1.13 1.16 1.19 1.20 1.21
where l is the length of the line; LC is the inductance value of the With CLR (p.u.) 1.11 1.15 1.18 1.21 1.22 1.23
CLR; le is the length of equivalent line of CLR, L0 and C0 are equiva- Rising rate (%) 0.91 1.77 1.72 1.68 1.67 1.65
lent inductance and capacitor of the transmission line. In most Single phase fault
cases, le is much smaller than l, so the quadratic term can be ignored No CLR (p.u.) 1.04 1.16 1.20 1.20 1.14.u. 1.04
With CLR (p.u.) 1.05 1.16 1.20 1.20 1.14 1.04
and the equation could be simplified as following:
Rising rate (%) 0.96 0 0 0 0 0
LC Load rejection
le ¼ ð5Þ No CLR (p.u.) 1.11 1.16 1.19 1.22 1.24 1.24
2L0
With CLR (p.u.) 1.13 1.18 1.22 1.25 1.27 1.27
Based on Eq. (5), in the simulations CLR can be substituted by a Rising rate (%) 1.80 1.72 2.52 2.46 2.42 2.42
segment of line with the length of le. In the sample case of this sec-
tion, the inductance value L0 of the transmission line is
0.08436 mH/km and LC is 25 mH thus the length of equivalent line According to Table 8, following conclusions can be drawn:
for the CLR is 14.82 km. The simulation results with the Equivalent
Circuit Method are shown in Table 7. (1) The CLR will lead to the power frequency and load rejection
As listed in Table 7, the results calculated by Equivalent Circuit overvoltages increase. The largest rising rates are 1.77% and
Method are close to the results of the original accurate method. 2.52% respectively. Overvoltages at the line terminal are
Equivalent Circuit Method can give an acceptable solution of much severer.
switching transient without CLR models. It can be used as a coarse (2) The single phase grounding fault overvoltages scarcely
approach to approximate the switching overvoltages. changes after CLRs installed.
(3) In the case studied in this section, with applied CLRs, the
4. Side effects of the CLR on temporary overvoltages severest temporary overvoltage is up to 1.27 p.u., extremely
close to the limit (1.3 p.u.).
Excessive sustained overvoltages may lead to damage of
transformers and other power system equipment. Temporary over-
5. Conclusions
voltages are a consequence of switching operations on long trans-
mission lines, or the switching of capacitive devices, and may
Installing CLR is a good choice to decrease the fault current lev-
result in arrester failures. Therefore the studies related to the tem-
els in the power system, whereas other problems will be caused at
porary overvoltages are necessary to ensure the safe operation [16].
the same time. Overvoltages of transmission line and equipment
The most common causes of temporary overvoltages are Ferranti
will rise up close to or even exceed the limit. Thus, attentions
effect, ground faults and load rejections. As shown in the previous
should be paid on the levels of overvoltages before CLR installed.
section, the CLR increases the equivalent inductance of transmis-
CLRs greatly increase the magnitude and shorten the rise-time
sion line and the temporary overvoltages may change as well
of the TRV, and consequently require limiting measures to restrain
[17]. In this section, the side effects of CLR on temporary overvol-
the much severer TRV. Installing parallel capacitors is widely
tages are discussed.
adopted to decrease TRV. If a CLR is applied, the capacitors
Taking an unload lossless transmission line for example, the ra-
installed across the reactor phases should be discussed further
tio of voltage at receiving end to voltage at sending end is [15]:
with the reactor manufacturer. In the case discussed above, 60 nF
U_ 2 cos u capacitor in the original plan cannot decrease TRV to the limit per-
K 12 ¼ ¼ ð6Þ fectly. Based on extensive researches, for this particular applica-
U_ 1 cosðal þ uÞ
tion, it was deemed appropriate to treat 1 lF and 0.8 lF parallel
pffiffiffiffiffiffiffiffiffiffi
where a ¼ x L0 C 0 is phase-shift
pffiffiffiffiffiffiffiffiffiffiffiffi coefficient, l is the length of line, capacitors at each end of the line respectively as a minimum capac-
u = arctan(XS/ZC), Z c ¼ L0 =C 0 is the line impedance and XS is the itance needed to control the TRV.
impedance of the CLR. For switching overvoltages and temporary overvoltages, install-
In Eq. (6), the CLR affects the overvoltage by u and the power ing CLR will make the overvoltages rise, except for the single phase
frequency overvoltage rises with CLR which lead to an increment fault overvoltage. Researches show that the overvoltages go up
of L. In the case that al + u = 90°, K12 tended to be infinite and with the increment of CLR impedance. In the case discussed in this
the overvoltages resonated. In order to make clear whether the article, the overvoltages are close to the limit with CLRs installed.
temporary overvoltages exceed the limit, related simulations The studies in this paper indicate that the application of CLR
are very necessary. The line parameters are the same to those in will lead to an increment of TRV, switching overvoltages and tem-
Table 3 and the results are listed in Table 8. porary overvoltages. These overvoltages may exceed the limit in
 S. Chen et al. / Electrical Power and Energy Systems 45 (2013) 340–345 345

some case, thus intensive simulations are strongly recommended [7] Harner RH, Rodriguez J. Transient recovery voltages associated with power
system, three-phase transformer secondary faults. IEEE Trans Power App Syst
before the CLR applied to ensure the overvoltages of equipment
1972;91(5):1887–96.
and transmission lines are well within the safe limits. [8] Hedman DE, Lambert SR. Power circuit breaker transient recovery voltages.
IEEE Trans Power App Syst 1976;95(1):197–207.
[9] Steurer M, Hribernik W, Brunke JH. Calculating the transient recovery voltage
References associated with clearing transformer determined faults by means of frequency
response analysis. IEEE Trans Power Del 2004;19(1):168–73.
[10] Swindler DL, Chwartz PS, Hamer PS, et al. Transient recovery voltage
[1] Amon FJ, Fernandez PC, Rose EH et al. Brazilian successful experience in the considerations in the application of medium-voltage circuit breakers. IEEE
usage of current limiting reactors for short circuit limitation. In: International Trans Ind Appl 1997;33(2):383–8.
Power Systems Transients (IPST’ 05). Montreal, Canada; 2005. [11] Wagner CL, Smith HM. Analysis of transient recovery voltage (TRV) rating
[2] Bellei TA, Camm EH, Ransom G. Current-limiting inductors used in capacitor concepts. IEEE Trans Power App Syst 1984;103(11):3354–63.
bank applications and their impact on fault current interruption. In: [12] Chen S, Xu W, Wang Z, et al. Effects of 500 kV current-limiting inductors on
Proceedings of the IEEE power engineering society transmission and transient recovery voltage and its countermeasures. North China Electr Power
distribution conference. vol. 1, 2001. p. 603–8. 2008;5(1):11–5.
[3] Peelo DF, Polovick GS, Sawada JH, et al. Mitigation of circuit breaker transient [13] DL/T 40222007. Order technical conditions of high voltage AC circuit breakers.
recovery voltages associated with current limiting reactors. IEEE Trans Power Beijing: China Electric Power Press; 2007.
Del 1996;11(2):865–71. [14] IEC 622712100. High-voltage alternation-current circuit-breakers, 2001.
[4] Kado H, Uriu Y, Ishigooka T. Line fault test of model power system with three- [15] Zhang Weibo, He Jinliang, Gao Yuming. Overvoltage protection and insulation
phase superconducting fault current-limiting reactors. Electr Eng Jpn coordination. Beijing: Tsinghua University Press; 2002.
2007;112(4):85–90. [16] Ketabi A, Ranjbar AM, Feuillet R. Analysis and control of temporary
[5] Smith RK. Tests show ability of vacuum circuit breaker to interrupt fast overvoltages for automated restoration planning. IEEE Trans Power Del
transient recovery voltage rates of rise of transformer secondary faults. IEEE 2002;17(4):1121–7.
Trans Power Del 1995;10(1):266–73. [17] Wang W, Chen S, Yang P. Effects of current-limiting inductor on power
[6] Montillet GF, Dufournet D. Transient recovery voltages requirements for frequency overvoltages in transmission line. Power Syst Technol
system source fault interrupting by small generator circuit breakers. IEEE 2010;34(3):193–6.
Trans Power Del 2002;17(2):474–8.