Professional Documents
Culture Documents
(4)
(5)
(6)
Fig. 2. Maximum peak values of a reactor’s transient overvoltage in relation
to VD for various values of VR.
(8) (12)
It can be easily observed, that the voltage expression (1) con- (13)
sists of two component groups, one with power frequency and
one with natural frequency . Similarly, the current expres- From (11)–(13), the following expression is obtained:
sion (2) consists of three component groups, one with power
(14)
frequency , one with natural frequency , and one dc com-
ponent. Therefore, the most adverse peak values of the current In most cases, shunt reactors are connected when the voltage
and voltage expressions equal to the sum of the peak values of rise tends to exceed 10% and the voltage compensation achieved
each respective component group are reaches even 100%. Therefore, for the range of values between
10% to 30% for VR and 0 to 100% for VD, respectively, the
(9) maximum peak value of the transient overvoltage varies ac-
cording to Fig. 2.
(10)
As can be easily seen in Fig. 2, even in cases the reactor
The maximum peak values for the transient overvoltage are is intended to compensate voltage drops in the order of 30%,
obtained for (k is any integer) (i.e., for closing the maximum transient overvoltage of the reactor is less than
at peak voltage between breaker poles). On the contrary, the 1.3 p.u. Therefore, the minimization of the maximum transient
maximum peak values for the inrush current are obtained for overvoltage may not be worth the application of controlled
(k is any integer) (i.e., for closing at zero voltage switching.
between breaker poles). The results are reversed for the respec- The maximum and minimum peak inrush current is derived
tive minimum peak values. Thus, at a first glance, the optimum from (10) for and , respectively
closing instant for the application of controlled switching is not
obvious as the minimization of transient overvoltage requires (15)
the maximization of inrush current and vice-versa.
The maximum transient overvoltage is derived from (9) for (16)
(18)
TSIREKIS et al.: CONTROL OF SHUNT REACTOR INRUSH CURRENTS 759
Fig. 3. Maximum peak values of reactor’s inrush current in relation to VD for Fig. 5. Contribution of natural frequency component to the total maximum
various values of VR. value of the reactor’s inrush current in relation to VD for various values of VR.
As shown in Fig. 5, this contribution is lower than 3% of the where ,, and are the closing instants of the three phases,
total maximum value of inrush current. Therefore, all shunt ca- the voltage across the breaker pole of the second
pacitances and series reactances (except, of course, that of the phase-to-close after the closing of the first phase, and and
760 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 20, NO. 2, APRIL 2005
Fig. 8. Diagram illustrating the making instant window for a case when the IV. CASE STUDIES
most favorable target instant corresponds to zero voltage across the breaker pole.
A. Shunt Reactors in the Hellenic Power System
HV shunt reactors are installed in six different locations in the
Hellenic-Interconnected Power System. Due to various reasons
related to the protection scheme used in the local substations,
CBs with independent pole operation are used in all cases. This
has led in some cases to phase and neutral overcurrents causing
unintentional tripping of phase and zero-sequence protection re-
lays. The application of controlled switching has been proposed
for the elimination of the high dc current components and an
investigation of its performance has been carried out for several
cases. One of these cases is presented in detail and the results of
all of the cases examined are summarized next in order to derive
general conclusions.
Although neutral is grounded in the actual reactor banks in-
stalled in the Hellenic-Interconnected System, for comparison
Fig. 9. Diagram illustrating the making instant window for a case when the purposes, investigation is carried out for both grounded and iso-
most favorable target instant corresponds to the peak voltage across the breaker lated neutral cases. The latter offer the most adverse conditions,
pole. as shown in Section II.
Another important characteristic of the shunt reactors exam-
which complicates the study of synchronized switching [1]–[5], ined is that their actual capacity is almost five times greater
[14], [16], [17]. In fact, in almost all cases, the instant of elec- than the conventional nominal capacity. This is due to several
trical closing (making instant) does not coincide with the in- changes in the HV system design during commissioning of the
stant of mechanical closing of the CB contacts (target instant). reactors. The practical consequence is that saturation does not
Making instant is determined by the intersection of the voltage occur during the operation of the reactors, even in the transient
waveform between the CB contacts and the dielectric strength period. Therefore, operation of the reactors at constant flux, as
characteristic of the contact gap. Statistical deviations in the assumed in Section II, is perfectly valid.
breaker operating time (the time interval between the signal and
the initiation of contact movement), the contact velocity, and the B. Simulation Tools
dielectric strength of the contact gap affect the target instant and The most adverse transients appearing during the uncon-
slope. These deviations result in random, parallel shifts to both trolled closing operation of the switching device are calculated
sides of the voltage withstand characteristic and deviations of its using the widely known ATP/EMTP computer program. The
slope. Thus, instead of a simple making instant and the respec- optimum switching instants are calculated by the controlled
tive target instant, it is more realistic to consider a “window” of switching calculation program (CSCP) developed at NTUA
possible making instants and the respective target instants, as il- [14]. Interaction of the three phases and the various statistical
lustrated in Figs. 8 and 9 [3], [16]. deviations of CB characteristics are considered in this program.
As shown in Section II, elimination of the dc components of
the inrush currents is achieved by closing at peak voltage across C. Typical System Configuration
breaker poles or for simultaneous closing of all three phases. Shunt reactor energization in the power system of the Greek
From the previous figures, it becomes obvious that the latter case island Cephalonia is used as a representative implementation of
is practically impossible due to the problem. This small power subsystem is fed by the intercon-
• statistical variation of the slope of dielectric strength char- nected power system of Greece through two lengthy HV subma-
acteristic; rine cables. For the absorption of the excessive reactive power,
762 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 20, NO. 2, APRIL 2005
Fig. 10. Network considered for shunt reactor energization. Black, empty,
and hatched boxes designate HV bus sections, feeders, and submarine cables,
respectively.
TABLE II
MAXIMUM p.u. INRUSH CURRENTS AFTER ENERGIZATION OF SHUNT
REACTORS WITH ISOLATED NEUTRAL ( 1T = 60 5
: ms, RDDS = 1 1 p.u.)
:
V. CONCLUSION
An implementation of synchronized switching to the ener-
gization of HV shunt reactors in the Hellenic-Interconnected
Power System has been presented. The most adverse conditions
for the generation of high energization stresses have been de-
scribed and the possible benefits obtained by means of syn-
chronized switching have been investigated. Various parame-
ters, such as neutral grounding condition, dielectric characteris-
tics, and statistical variations of the switching device, affecting
the effectiveness of this modern technique, have been taken into
account. It can be summarized that the main requirement for ef-
fective application of synchronized switching is the use of con-
trolled switching arrangements with quite a small deviation of
starting instant of contacts movement and high contact speed.
With this requirement fulfilled, a very sufficient reduction of
Fig. 14. Effect of RDDS and 1T to the maximum inrush current of the third the dc current components appearing after the energizations of
phase-to-close upon controlled energization of the reactor with isolated neutral.
shunt reactors is achieved, but only for grounded neutral cases.
For isolated neutral cases, the requirements are quite stringent,
TABLE I making practically controlled switching applications nonbene-
MAXIMUM p.u. PHASE AND NEUTRAL INRUSH CURRENTS AFTER
ENERGIZATION OF SHUNT REACTORS WITH GROUNDED NEUTRAL
ficial.
REFERENCES
[1] “Controlled switching of HVAC circuit-breakers—planning, specifica-
tion and testing of controlled switching systems,” Electra, no. 197, pp.
23–33, Aug. 2001.
[2] U. Krüsi and K. Fröhlich, “Model based determination of circuit breaker
characteristic for controlled switching,” in Proc. IASTED Int. Conf.
Modeling, Identification Control, Innsbruck, Austria, Feb. 2001, paper
no. 324-131.
[3] “Controlled switching of HVAC circuit-breakers—guide for application
lines, reactors, capacitors, transformers,” Electra, pt. 1st, no. 183, Apr.
1999.
[4] R. Rocha and J. L. Tavora, “EMTP model for controlled switching by
means of a TACS routine,” in Proc. Int. Conf. Power Systems Transients,
achieves a reduction of inrush currents below 1.5 p.u., thus Seattle, WA, Jun. 1997, pp. 254–259.
avoiding undesired tripping of phase- and zero-sequence pro- [5] “Controlled switching: a state-of-the-art survey,” Electra, pt. 1st, no.
tection relays. 163, Dec. 1995.
[6] A. H. Khan, D. S. Johnson, J. H. Brunke, and D. L. Goldsworthy, “Syn-
Similarly, for reactors with isolated neutral (Table II), the chronous closing applications in utility transmission systems,” in Proc.
values 0.5 ms and 1.1 p.u. are considered for and 36th CIGRE Session, Paris, France, 1996, paper 13-306.
764 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 20, NO. 2, APRIL 2005
[7] A. C. Carvahlo, W. Hofbauer, P. Högg, and K. Fröhlich, “Controlled Constantinos D. Tsirekis was born in Thessaloniki, Greece, in 1969. He re-
switching as a reliable mean to reduce stresses imposed to the circuit- ceived the Diploma in electrical engineering from the National Technical Uni-
breaker and to the network,” in CIGRE SC 13 Colloquium, Florianópolis, versity of Athens (NTUA), Athens, Greece, in 1994.
Sep. 1995. Currently, he is with the Implementation Planning Department of the Hellenic
[8] J. Bachiller and J. Sanchez, “Controlled switching—elimination of Transmission System Operator. He is also a Postgraduate Student at NTUA. His
transients in the operation of shunt reactors,” in Proc. IEEE Int. Symp. research interests include switching transients analysis and control.
Electric Power Engineering, vol. 6, Stockholm, Sweden, Jun. 1995, pp.
158–162.
[9] A. Holm, R. Alvinsson, U. Akesson, and O. Karlen, “Development of
controlled switching of reactors, capacitors, transformers and lines,” in
Proc. 33rd CIGRE Session, Paris, France, 1990, paper 13-201.
[10] H. Suiter, W. Degen, H. Eggert, and H. H. Schramm, “Consequences of
controlled switching for system operation and circuit-breaker behavior,” Nikos D. Hatziargyriou (SM’90) was born in Athens, Greece, in 1954. He re-
in Proc. 33rd CIGRE Session, Paris, France, 1990, paper 13-202. ceived the Diploma in electrical and mechanical engineering from the National
[11] B. J. Ware, J. G. Reckleff, G. Mauthe, and G. Schett, “Synchronous Technical University of Athens (NTUA), Athens, Greece, in 1976, and the M.Sc.
switching of power systems,” in Proc. of the 33rd CIGRE Session, Paris, and Ph.D. degrees from University of Manchester Institute of Science and Tech-
France, 1990, paper 13-205. nology (UMIST), Manchester, U.K., in 1979 and 1982, respectively.
[12] A. Andersen, S. Berneryd, and S. Lindahl, “Synchronous energization Currently, he is Professor in the Electrical and Computer Engineering Depart-
of shunt reactors and shunt capacitors,” in Proc. 32nd CIGRE Session, ment of NTUA. His research interests include modeling and digital techniques
Paris, France, Aug./Sep. 1988, paper 13-12. for power system analysis and control.
[13] R. J. Rajotte, C. Charpentier, S. Brault, H. H. Le, and H. Huny, “Field Dr. Hatziargyriou is currently Chairman of the Greek Power Chapter of the
tests of a circuit-breaker synchronous control,” in Proc. IEEE Power IEEE, and member of CIGRE SC38 and of the Technical Chamber of Greece.
Eng. Soc. Winter Meeting, vol. 13, Apr. 1995, pp. 472–479.
[14] C. D. Tsirekis, N. D. Hatziargyriou, and B. C. Papadias, “Controlled
switching based on the injection method,” in Proc. Int. Conf. Power Sys-
tems Transients, vol. II, Rio de Janeiro, Brazil, Jun. 2001, pp. 405–410.
[15] W. M. C. Van Den Heuvel and B. C. Papadias, “Interaction between
phases in three-phase reactor switching—II. Ungrounded reactors,”
Electra, pt. 1st, no. 91, pp. 11–50, Dec. 1983. Basil C. Papadias was born in Greece in 1932. He received the Electrical Engi-
[16] “Controlled switching of HVAC circuit-breakers—guide for application neering and Doctor of Engineering degrees from the National Technical Univer-
lines, reactors, capacitors, transformers,” Electra, pt. 2nd, no. 185, Aug. sity of Athens (NTUA), Greece, in 1956 and 1969, respectively, and the Masters
1999. and Doctor of Engineering degrees from Ransselaer Polytechnic Institute, Troy,
[17] “Controlled switching: a state-of-the-art survey,” Electra, pt. 2nd, no. NY, in 1972 and 1975, respectively.
164, Feb. 1996. From 1958 to 1975, he was with Public Power Corporation, Athens, Greece.
[18] W. M. C. Van Den Heuvel and B. C. Papadias, “Interaction between Since 1975, he had been Professor and Director of the Electrical Energy Systems
phases in three-phase reactor switching—II. Ungrounded reactors,” Lab of the Department of Electrical and Computer Engineering of NTUA. Prof.
Electra, pt. 2nd, no. 81, pp. 57–81, May 1987. Papadias passed away on April 4, 2002.