Electric Power Systems Research 116 (2014) 419–426

Contents lists available at ScienceDirect

Electric Power Systems Research

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

Zero sequence behaviour of a double-circuit overhead line

Roberto Benato a,∗ , Sebastian Dambone Sessa a , Fabio Guglielmi a ,
Ertugrul Partal b , Nasser Tleis c
a
Department of Industrial Engineering, University of Padova, Via Gradenigo, 6/A, 35131 Padova, Italy
b
National Grid Electricity Transmission, Gallows Hill, Warwick CV34 6DA, United Kingdom
c
Power & Water Planning Division, P.O. Box 564, Dubai, United Arab Emirates

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

Article history: Positive, negative and zero sequence series impedances (in /km units) are the first step for power flow
Received 19 November 2013 and short circuit analyses and are also necessary for setting the distance relays. The paper is devoted to
Received in revised form 11 June 2014 the computation of zero sequence series impedance (in /km units) (self and mutual) in double-circuit
Accepted 13 July 2014
overhead lines (OHLs) with one or more earth wires. A systematic comparison with matrix approaches
demonstrates that IEC 60909-2 formula adds the zero sequence self and the zero sequence mutual
Keywords:
impedance between the two circuits. This approximation of having a unique value of zero sequence
Sequence impedances
impedance is misleading and gives great errors when phase-to-ground short circuit occurs along a cir-
Double-circuit OHL
Multiconductor Cell Analysis
cuit. It is worth noting that the zero sequence mutual impedance between parallel circuits installed on
Clarke’s equivalent circuit the same tower may be as high as 50% (or more) of the zero sequence self impedance of a circuit alone.
Consequently, zero sequence mutual impedance between the two circuits cannot be neglected or mixed.
MCA (Multiconductor Cell Analysis) shows that the simplified IEC approach underestimates heavily (more
than 18%, depending upon the spacing between the circuits) the phase-to-ground fault current for a fault
occurrence along one circuit of the double circuit.
© 2014 Published by Elsevier B.V.

1. Introduction different from the zero sequence self impedance. Therefore it is

fundamental to consider them separately for the computations
The selectivity of the protection system in a meshed network of all those short circuits involving the earth. The paper shows a
strictly depends upon the correct settings of the distance relays. comparison of zero sequence impedances computed by means of
As it is well known, their settings are based on positive (nega- two approaches: the IEC 60909-2 [7] one and a matrix one which
tive) and zero sequence series impedances [1,2]. Moreover, unless starts from Carson impedance matrix. The different zero sequence
in-depth multiconductor short circuit computations are necessary impedances are used to compute the effect on a phase-to-ground
(e.g. knowledge of the current sharing between phase conductors, short circuit occurring in different location along one circuit of the
earth wires and the ground [3]), short circuit current magnitudes double-circuit. In conclusion, the two aforementioned approaches
are usually computed by means of the sequence theory. Conse- are compared with the most accurate MCA procedures without
quently, suitable and accurate formulae of kilometric sequence simplifications or approximations.
impedances must be used. For EHV cable lines, previous papers
have been published [4–6]. This paper is devoted to the zero
2. Zero sequence behaviour of a double-circuit overhead
sequence series impedance for a double-circuit OHL and to the
lines
short circuit (involving the earth) computations. In a double-circuit
OHL, the influence of the zero sequence mutual impedance on the
Let us consider a typical UK un-transposed double-circuit EHV
phase-to-ground short circuit current, occurring along the line, is
overhead line shown in Fig. 1 [8]. The conductors and earth wire
characteristics are reported in Table 1. This table also reports
the positive sequence parameters of the double-circuit OHL. It
∗ Corresponding author. Tel.: +39 0498277532; fax: +39 0498277599. is worth noting that the double-circuit OHL length is equal to
E-mail addresses: [email protected], [email protected] 135.42 km which is the maximum 400 kV OHL length in UK
(R. Benato). (i.e. Cottam-Eaton Socon OHL). For the sake of completeness, in

http://dx.doi.org/10.1016/j.epsr.2014.07.011
0378-7796/© 2014 Published by Elsevier B.V.
420 R. Benato et al. / Electric Power Systems Research 116 (2014) 419–426

Table 1
Nomenclature Data assumed in the MCA of 400 kV double-circuit OHL in UK grid.

Line length d km 135.42

z- 0 ca series self impedance of one circuit (e.g. circuit a) Span length km 0.366
z- 0 a b series mutual impedance between circuit a and cir- Earth resistivity m 20
cuit b Substation earthing  0.1
resistance
OHL overhead lines
Tower earthing  10
MCA Multiconductor Cell Analysis resistance
HV high voltage Bundled Conductor 4 sub-cond.
EHV extra high voltage ACSR Zebra
EW earth wire ˚ = 28.62 mm
Spacing = 0.3 m
TSO Transmission System Operator
Earth wire (EW) ACSR Zebra
ACSR Aluminium Conductor Steel Reinforced ˚ = 28.62 mm
ST sequence theory Positive sequence m/km 8.25
 distance from sending end substation series resistance at
20 ◦ C (50 Hz)
 distance from receiving end substation
Positive sequence mH/km 0.422
d line length series inductance
Sc fault section Positive sequence nS/km 33.56
shunt leakance
Subscripts (50 Hz)
Positive sequence ␮F/km 0.0267
REC. receiving
capacitance
SEND. sending Single circuit ampacity A 4 × 800 = 3200
SUB SEND. substation sending Ia referred to winter
SUB REC. substation receiving rating [8]
Winter rating [8] of MVA 2 × 2217
double-circuit OHL

means of the simplified Carson-Clem [9] or complete Carson

Table 1, the ampacity and winter rating are also reported but
theory [10]:
they are useless in short circuit computations since the network
must be at no-load (shunt admittances can be neglected after IEC Z tot
-
60909).  
 zph11 z ph12 z ph13 z ph14 z ph15 z ph16 z pe17 
The resistance of the bundle conductor and of the earth wire  
must be computed at 20 ◦ C since, according to IEC 60909, the con- z 
 ph21 z ph22 z ph23 z ph24 z ph25 z ph26 z pe27 
ductors must be “cold” for maximum short circuit computations  
z z pe37 
◦ =
rPC 20◦ C
=
0.066
= 0.0165 /km;  ph31 z ph32 z ph33 z ph34 z ph35 z ph36
r
i.e. ph 20 C  
n 4  zph41 z ph42 z ph43 z ph44 z ph45 z ph46 z pe47  (1)
rew 20◦ C = 0.066 /km.  
 
In the following, a matrix technique which starts from the  zph51 z ph52 z ph53 z ph54 z ph55 z ph56 z pe57 
kilometric impedance matrix of all the conductors (active and  
z 
passive ones) is proposed. Fig. 2 shows the circuital scheme in  ph61 z ph62 z ph63 z ph64 z ph65 z ph66 z pe67 
 
order to model and to compute the zero sequence impedances z z ep72 z ep73 z ep74 z ep75 z ep76 z 
ep71 ew77
(both self and mutual ones) of a double-circuit OHL with one
earth wire. First the impedance matrix Ztot of all 7 conduc- In the hypothesis that earth wire voltage is constantly zero, the
tors (phase conductors plus earth wire) can be obtained by “elimination technique of the grounded conductors” [11] can be

7
IEW
0.3 7, EW 0.3
1
I0
6.93 m 6.93 m I0 2
0.3 0.3 I0 3
1, R 4, T I0 circuit a
10.16 m 10.16 m I0
2, S 0.3 0.3 5, S I0 4

8.33 m 8.33 m ht=50.04 m 6

3, T 6, R circuit b
U04 U05 U06
PH sag=11.02 m ht=29.14 m
ht=39.61 m
ht=20.30 m
EW sag=10.42 m
3I0-IEW

Fig. 1. National Grid UK typical double-circuit overhead line (not to scale) with
asymmetric phasing (RST-TSR also called low-reactance phasing). Fig. 2. Zero sequence scheme of a double circuit OHL with one earth wire.
 R. Benato et al. / Electric Power Systems Research 116 (2014) 419–426 421

applied. This yields the reduced matrix Zph red which includes the It is possible to demonstrate that Eqs. (2) and (3) can be re-
effect of the earth wire: written as:
Z ph red
   z 2ca ew
z0 ca = z0 − 3 · ˝/km (4)
 z11 z 12 z 13 z 14 z 15 z 16  z ew
 
z 
 21 z 22 z 23 z 24 z 25 z 26  z 2ca ew
  z0 = 3 · zm −3· ˝/km (5)
z z 36 
a b ca cb
z ew
 31 z 32 z 33 z 34 z 35
 
 z41 z 42 z 43 z 44 z 45 z 46  where z0 = z11 + z12 + z13 [/km] is the zero sequence impedance of
 
  one circuit; zca ew [/km] is the mutual impedance between circuit
 z51 z 52 z 53 z 54 z 55 z 56 
  a (phases 1, 2 and 3) and the earth wire ew; zew [/km] is the self
z z 62 z 63 z 64 z 65 z
 impedance of earth wire ew; zm ca cb [/km] is the average mutual
61 66
impedance between the phases of circuits a and b.
By applying three equal current generators I0 [12] to one circuit If in Fig. 2 the zero sequence current generators are imposed at
of the double-circuit it is possible, after some passages, to compute both circuits, there is a mix of zero sequence self impedances with
both the zero sequence self impedance of the supplied circuit and zero sequence mutual ones namely:
the mutual one between the two circuits by means of:
Z ph red

     
Z ph red

    
 U 01   z11 z 16  I 
z 16   I 
z 12 z 13 z 14 z 15
 U 01   z11 z 12 z 13 z 14 z 15      0
     0 U  z   
U  z 
z 26   I   02   21 z 22 z 23 z 24 z 25 z 26   I 0 
 02   21 z 22 z 23 z 24 z 25      
     0  U  z z 36   I 0 
U  z z 36   I   03   31 z 32 z 33 z 34 z 35
 03   31 z 32 z 33 z 34 z 35  = · 
 =  ·  0   U 04   z41 z 42 z 43 z 44 z 45 z 46   I 0 
 U 04   z41 z 42 z 43 z 44 z 45 z 46   0       
           
     U 05   z51 z 56   I 0 
z 56   0 
z 52 z 53 z 54 z 55
 U 05   z51 z 52 z 53 z 54 z 55      
      U  z  I 
U  z z 62 z 63 z 64 z 65 z
 0 06 61 z 62 z 63 z 64 z 65 z66 0
06 61 66
 
     z01 + z0m1 cb 
 z11 + z12 + z13   z01   
    z +z 
z +z +z   z   02 0m2 cb 
 21 22 23   02   
    z +z 
z +z +z   z   03 0m3 cb 
 31 32 33   03  = ·I (6)
=  · I0 =  ·I  z0m4 ca + z04  0
 z41 + z42 + z43   z0m4 ca  0  
     
     z0m5 ca + z05 
 z51 + z52 + z53   z0m5 ca   
    z 
z +z +z  z  0m6 ca+z 06
61 62 63 0m6 ca

The three zero sequence self impedances are different for each Therefore in the (6 × 1) column of the matrix in (6), it is impos-
conductor 1, 2 and 3 and are given by: sible to have a clear distinction between the zero sequence self
and the mutual impedances. The average value of the six elements
U 01 U 02 U 03 inside the column can be expressed as:
z 01 = ; z 02 = ; z 03 =
I0 I0 I0
In order to have a unique zero sequence self impedance of the z 2ca ew
z 0 av = z 0 + 3 · z m ca cb −6· (7)
circuit, this average value can be computed: z ew

z 01 + z 02 + z 03 It is immediate to verify that formula (7) is equal to the addition

z0 ca = . (2)
3 of (4) and (5). In order to compute the zero sequence impedance
From the same scheme of Fig. 2, it is also possible to obtain per circuit in a double-circuit OHL with one earth wire, IEC 60909-2
the zero sequence mutual impedance between the two circuits [7] gives the following formula:
as an average value of the three different zero sequence mutual
impedances namely: z  2QLE
z  0−IEC = z  0 + 3 · z  LME − 6 · (8)
z  QQE
U U U
z 0m4 ca = 04 ; z 0m5 ca = 05 ; z 0m6 ca = 06 ,
I0 I0 I0
Even with different symbols, IEC formula (8) is equal to (7)
and consequently derives from supplying both circuits with zero
z 0m4 ca + z 0m5 ca + z 0m6 ca
z0 b a = . (3) sequence current generators and hence adds the zero sequence
3
self and zero sequence mutual impedances. In Appendix A, details
If the two circuits have the same geometrical configuration with of the aforementioned procedure are given. In technical literature
respect to the tower axis, then it is: there are some Italian contributions to compute single circuit zero
sequence impedance taking footing tower resistances and substa-
z0 cb = z0 ca and z 0 a b = z0 b a tion earthing grid ones into account [13]. In fact, the regime of
This method gives the possibility to correctly compute the induced currents in the earth wire are slightly influenced by its
phase-to-ground short circuit current at any point along the line footing tower resistances and substation earthing grid ones: these
including both ends. induced currents influence the series zero sequence impedance.
422 R. Benato et al. / Electric Power Systems Research 116 (2014) 419–426

DOUBLE-CIRCUIT The well-known following expression [15] can be used to com-

Substation OHL Substation
pute the phase-to-ground fault current I1p :
SEND. REC.
1.1 · 3 · E v
I3-ph = x1 kA I3-ph = x2 kA I 1p = (9)
Sc Z1 + Z2 + Z0
I1-ph = y1 kA ' I1-ph = y2 kA
where Z0 [] is the zero sequence impedance as seen from the
fault location; Z1 = Z2 [] are the positive and negative sequence
d impedances as seen from the fault location.
The computation of sequence impedances as seen from the fault
Fig. 3. Phase-to-ground short circuit occurring at  distance from sending-end. location can be derived by making use of delta-wye transforma-
tions.
3. Phase-to-ground short circuit calculation occurring In order to demonstrate that IEC approach is only correct if fault
along one circuit of the double-circuit OHL location is at sending or receiving ends it is possible to apply the
aforementioned procedures in these cases. Eq. (10) represents the
Let us suppose a phase-to-earth short circuit occurring at a dis- zero sequence impedance as seen from fault location at sending
tance  from sending-end ( = d −  from receiving-end) as shown end.
 
in Fig. 3. The double-circuit is in electrical parallel. It is worth Z 0 ca +Z 0 a b
Z 0sub SE · 2
+ Z 0sub RE
remembering that in order to characterize the end substations it Z0 =   (10)
is necessary to know the fault levels (three-phase x1 , x2 and single- Z 0 ca +Z 0 a b
Z 0sub SE + 2
+ Z 0sub RE
phase y1 , y2 sub-transient short circuit currents usually given by
Transmission System Operator (TSO)). The zero sequence single- where the part concerning the double circuit in parallel is simply
phase circuit is depicted in Fig. 4 [14]. The meaning of the symbols
is cleared in Appendix B. The total impedances are derived from the Z0 ca + Z0 a b Z 0 IEC
= .
kilometric ones as in the following 2 2
In order to consider the OHL system in its real asymmetry,
Z0 ca = z0 ca ·  [˝], Z- 0 ca = z0 ca ·  [˝]
without simplified and approximated hypotheses, the short circuit
occurrence must be also investigated by means of MCA.
Z0 cb = z0 ca ·  [˝], Z- 0 cb = z0 cb ·  [˝]
3.1. Case-study
Z0 a b = z0 a b ·  [˝], Z- 0 a b = z0 a b ·  [˝]
In this section the proposed procedures are applied to the
In order to calculate the phase-to-ground short circuit current double-circuit OHL 400 kV 4 mm × 400 mm ACSR (quadruple bun-
in the situation of Fig. 4, it is fundamental to consider the different dle conductor), whose electrical and geometrical characteristics are
role played by the zero sequence self and mutual impedances so to reported in Table 1. The tower arrangement is shown in Fig. 1.
derive the equivalent circuit (see Appendix B) of Fig. 5. Let us consider a phase-to-ground short circuit occurring at mid-
In Fig. 5, the values of impedances Za , Z a , Zb , Z b , can be derived line (i.e. at 67.71 km from both ends, see Fig. 6). In order to calculate
as in the following the sequence impedances, it is necessary (as already explained in
Za = Z0 − Z0 [˝], Z- a = Z- 0 − Z- 0 [˝] Section 2) to calculate the matrix Ztot by means of the complete
ca a b ca a b
Carson formulae (see Appendix A).
From Zseq matrix of Appendix A, the values of zero sequence are:
Zb = Z0 cb − Z0 a b [˝], Z- b = Z- 0 cb − Z- 0 a b [˝]
Z0 ca = Z0 cb = 0.1084 + j0.7832 [˝/km] (11)
' Z0 a b = Z0 b a = 0.0914 + j0.4153 [˝/km] (12)

The value of zero sequence impedance after IEC 60909-2 is given

z0 _ a _ b z 0 _ ca z 0 _ ca ' by:
z0 _ a _ b '
z- 0−IEC = 0.2019 + j1.2009 [˝/km] (13)
Sc
Once again, it is worth nothing that it is the sum of Z0 ca and
Z 0 sub _ SE z 0 _ cb z 0 _ cb ' Z 0 sub _ RE Z0 a b in (11) and (12). It is possible to compute the phase-to-ground
short circuit current by means of sequence theory (ST) and of MCA.
In order to highlight the misleading nature of IEC zero sequence
Sub. sending Sub. receiving
impedance, the ST is applied with IEC formula itself and with (zero
Fig. 4. Zero sequence circuit of phase-to-ground short circuit occurring at a distance sequence self and mutual impedances) equivalent wye circuit. The
 from sending-end. comparison between the three different approaches is shown in
Table 2. In case of short circuit occurring along one circuit of the
Z0_a_b Za Z'a Z'0_a_b
Substation DOUBLE-CIRCUIT Substation
OHL
Sc SEND. REC.
I3-ph =42.62 kA Sc I3-ph =32.27 kA
Z 0 sub _ SE Z 0 sub _ RE
Zb+Z'b I1-ph =43.42 kA 67.71 km 67.71 km I1-ph =28.27 kA

Sub. sending Sub. receving d=135.42 km

Fig. 5. Zero sequence circuit with “manifested” mutual impedances. Fig. 6. Phase-to-ground short circuit occurring at midline.
 R. Benato et al. / Electric Power Systems Research 116 (2014) 419–426 423

Table 2
Comparison values.

I1p at the midline [kA] I1p at sending-end [kA] I1p at receiving-end [kA]

R-ground 13.791 50.833 35.032

MCA S-ground 13.183 50.873 34.957
T-ground 12.952 50.833 35.032

Average MCA 13.309 50.846 35.007

ST with equivalent wye circuit 13.475 50.799 34.948
ST with IEC formula 10.870 50.784 34.933
Percentage difference equiv. wye circuit with MCA 1.24% −0.09% −0.16%
Percentage difference IEC with MCA −18.32% −0.12% −0.21%

Fig. 8. Measurement of double-circuit overhead line phase impedances with earth

return (the two circuits are symmetrical with respect to the tower).

Fig. 7. Behaviours of phase-to-ground short circuit, phase current magnitudes along

OHL. measurements on several double-circuit 400 kV, 275 kV and
132 kV OHLs. The objective was to measure the self impedances
double-circuit, IEC formula, used in ST, gives very high underesti- and mutual impedances, all with earth return, and with the line
mates (up to −18.32%). in actual service conditions. The measured impedances were
Fig. 7 (obtained with MCA) shows the phase-R-to-ground short then used to construct the 6 × 6 phase impedance matrix of the
circuit current magnitudes in the phases of faulted and unfaulted line and obtain the positive sequence impedance per circuit,
circuit. The only non-negligible short circuit current flows in 1-R the zero sequence impedance per circuit and the zero sequence
and in the phase which is in electric parallel with it, i.e. 6-R [3]. mutual impedance between the two circuits. Fig. 8 illustrates the
measurement approach.
3.2. General considerations in practical networks
U1 U2 U3
Z 11 = , Z 12 = Z 21 = , Z 13 = Z 31 = ,
I1 I1 I1
In practical networks, there are many double-circuit OHLs (e.g.
99% of the National Grid Electricity Transmission system consists U4 U5 U6
Z 14 = Z 41 = , Z 15 = Z 51 = , Z 16 = Z 61 =
of double-circuit OHLs). Whereas some of these lines are bussed I1 I1 I1
at both ends (as in the case study), others are only bussed at one
end. Therefore, depending on the fault location, there will be some By applying the voltage generator U to each phase in turn, the
double-circuit lines where currents in each circuit flow in the same complete range of self and mutual impedances may be measured.
direction, and other double-circuit lines where currents will flow in By assuming a transposed line, the self and mutual impedances per
opposite directions. The net effect on the earth fault current magni- (Z +Z +Z ) (Z +Z +Z )
circuit are given by:Z S = 11 322 33 and Z M = 12 313 23
tude depends on the balance of these effects as well as the proximity The positive sequence and the zero sequence impedances per
of a particular type of double-circuit line to the fault location. On circuit are given by:
the England & Wales transmission system, it has been found that
the effect of the zero sequence mutual impedance on earth fault (Z 11 + Z 22 + Z 33 ) − (Z 12 + Z 13 + Z 23 )
Z1 =
current magnitudes ranges from around 5% to 12%. Consequently, 3
in practical networks, the zero sequence mutual impedance should
not be lumped with the zero sequence self impedance. As through- (Z 11 + Z 22 + Z 33 ) + 2(Z 12 + Z 13 + Z 23 )
Z0 =
out demonstrated in the paper, the correct approach is to consider 3
the former as a separate item and build the network admittance The zero sequence mutual impedance between the two circuits
matrices accordingly. is given by (3) and here repeated:
Z 14 + Z 15 + Z 16 + Z 24 + Z 25 + Z 26 + Z 34 + Z 35 + Z 36
4. Comparison with U.K. measurements on double-circuit Z 0M =
OHL 3

4.1. Measurement technique 4.2. Measurement results

Over a period of many years, the Central Electricity Generat- The measured impedances for a double-circuit 400 kV line are
ing Board (CEGB) in the UK, the predecessor of National Grid summarized in Table 3, and compared with the calculated values
Electricity Transmission (NGET), carried out impedance on the base of data in Table 1.
424 R. Benato et al. / Electric Power Systems Research 116 (2014) 419–426

Table 3
A
Comparison between measured and calculated values.
Zaa Za= Zaa - Zab
Parameter (/km) 400 kV double-circuit OHL, 4 mm × 400 mm A
C Zc=Zab
Measured Calculated
Zab
C
zS 0.041 + j0.424 0.0474 + j0.4466
zM 0.023 + j0.148 0.0305 + j0.1683 Zbb B Zb= Zbb - Zab
z1 0.018 + j0.276 0.0168 + j0.2783
z0 0.087 + j0.720 0.1084 + j0.7832 1) B 2)
z0M 0.079 + j0.380 0.0914 + j0.4153

Fig. B1. Transformation from circuit (1) with “embedded” mutual impedance to
circuit (2) with “manifested” mutual impedance.

The accuracy of the measured values of resistance depends

greatly on the accuracy of measurement of the current phase angle Appendix A. Fortescue transformation matrices
and it is likely that some error was present in the phase angle mea-
suring equipment. This would account for the discrepancies in the By starting from the total conductor impedance matrix Ztot the
resistance values. However, the values of reactance are less suscep- reduced impedance matrix Zph red can be obtained by means of
tible to such an error and in fact close agreement exists between well-known elimination technique [11] i.e.:
the calculated and measured values of X1 . The lower measured zero
sequence reactance is probably due to a low value of earth resisti- Z ph red = Z PP − Z P EW · Z −1
EW · Z EW P
vity (hypothesized in computed values  = 20  m) in the vicinity
where ZPP is the submatrix of Ztot from row 1 to 6 and column from
of the line route but is still within the expected ±10% bound. More-
1 to 6, ZP EW the submatrix of Ztot row from 1 to 6 and column 7,
over also the assumption of symmetric line (or transposed one)
ZEW P the submatrix of Ztot row 7 and column from 1 to 6 and ZEW
gives slight errors.
is the element (7,7) of Ztot .
The sequence matrix can be easily derived by means of the for-
5. Conclusions
mula [10]:
Overhead lines are the backbone of any electric energy HV–EHV Z S = F · Z ph red · F −1
transmission grid.
Consequently, reliable and precise formulae for sequence where F is given by:
impedances are very important in power system modelling and  
1 1 1 0 0 0
computation. The aim of this paper is that, for a double-circuit  1 ˛ ˛2 0 0 
OHL, the zero sequence mutual impedance should not be implic-  0
1  1 ˛2 ˛ 0 0 0 
itly included in the zero sequence self impedance as IEC 60909-2 F = ·
3 0 0 0 ˛ ˛2 1
assumes. This IEC assumption can give errors up to 20% in the  
 0 0 0 ˛2 ˛ 1
phase-to-ground short circuit current when the fault occurs along 0 0 0 1 1 1

the line. The paper shows both a matrix computation of zero
sequence impedances of a double-circuit OHL and a comparison For the case-study of Fig. 1 the following sequence matrix can
of phase-to-ground short circuit currents with correct wye circuit be obtained as in the bottom of the page.
along with MCA. The paper originality is given by the application of The matrix ZS is full since the system is asymmetric. If it had
MCA to double-circuit OHL and of a simple but correct single-phase been symmetric (structurally or by means of phase transpositions)
model of a double-circuit OHL including mutual zero sequence the matrix ZS would have been almost diagonal with only two off-
impedance. diagonal elements Z0 a b and Z0 b a . It is worth noting that the self
positive and negative sequence impedances are equal, the mutual
impedances between sequences are all negligible except that of
zero sequence between the circuit a and b.
 R. Benato et al. / Electric Power Systems Research 116 (2014) 419–426 425

Fig. C1. Partitioned matrix of i = Y TOT · u in short circuit occurrence.

Appendix B. Equivalent circuits taking zero sequence Solving this system, it is possible to obtain the values of Za , Zb
mutual impedance into account and Zc of the equivalent wye circuit:

By definition, zero sequence currents I0 are equal in magni- Z a = Z aa − Z ab

tude and angle. Zero sequence mutual impedances are therefore Z b = Z bb − Z ab
independent of the phasing arrangement of the two circuits (low Z c = Z ab
reactance or super bundle phasing). The circuit consisting of two
self impedances Zaa and Zbb connected at one end, with mutual
impedances Zab between them depicted in Fig. B1(1) can be trans- Appendix C. Brief recalls of MCA short-circuit analysis [4]
formed in the equivalent wye circuit of Fig. B1(2).
The branches Za , Zb and Zc of the equivalent wye are determined By considering the fault levels, the sequence impedances for
by applying a voltage between two terminals when the third is each substation can be computed by the well-known formulae
open. The impedance between terminals A and C in the given circuit [3,15]:
with terminal B open is Zaa , and in the equivalent circuit Za + Zc .
E PV
Similarly, the impedance between terminals B and C with A open I  3P = 1.1 · (C.1)
is Zbb in the given circuit and Zb + Zc in the equivalent circuit. With Z1
a voltage U applied between terminals A and B with C open, in 3E PV
the given circuit the current I enters at A and leaves at B, flowing I  1P = 1.1 · (C.2)
Z  1 + Z 2 + Z 0
in opposite direction through the self-impedances Zaa and Zbb in
series. The voltage drop between terminals A and B is therefore By solving the Eqs. (C.1) and (C.2) it yields:
given by:
E PV
Z  1 = Z 2 = 1.1 · (C.3)
U = Z aa I − Z ab I + Z bb I − Z ab I = (Z aa + Z bb − 2Z ab )I I  3P

3 2
In the equivalent circuit the impedance is Za + Zb . Z 0 = 1.1 · E PV · − (C.4)
I  1P I  3P
The three equations expressing Za , Zb and Zc in terms of Zaa , Zbb
and Zab are: Once the diagonal sequence matrix ZS has been computed, it
is easy to compute the phase matrix Zpha by means of Fortescue
Z aa = Z a + Z c matrix F (as clearly shown in Appendix A). Once the YTOT has been
Z bb = Z b + Z c computed the faulty regime can be evaluated with a suitable par-
Z aa + Z bb − 2Z ab = Z a + Z b tition of it as in Fig. C1.
426 R. Benato et al. / Electric Power Systems Research 116 (2014) 419–426

The study of short circuit leads to the following equations: [4] R. Benato, Multiconductor analysis of underground power transmission sys-
tems: EHV AC cables, Elect. Power Syst. Res. 79 (January (1)) (2009) 27–38,
iphaS = Y 1 uphaS + Y 2 ux + Y 3 uphaR (C.5) http://dx.doi.org/10.1016/j.epsr.2008.05.016.
[5] R. Benato, L. Caciolli, Sequence impedances of insulated cables:
0 = Y 4 uphaS + Y 5 ux + Y 6 uphaR (C.6) measurements versus computations, in: Transmission and Distribu-
tion Conference and Exposition (T&D), IEEE PES, Orlando, FL, 2012,
iphaR = Y 7 uphaS + Y 8 ux + Y 9 uphaR (C.7) http://dx.doi.org/10.1109/TDC.2012.6281574.
[6] R. Benato, M. Forzan, M. Marelli, A. Orini, E. Zaccone, Harmonic
where the vector ix is null (see Fig. C1). From Eq. (C.6), it is rather behaviour of HVDC cables, Elect. Power Syst. Res. 89 (2012) 215–222,
immediate to obtain the unknown vector ux : http://dx.doi.org/10.1016/j.epsr.2012.03.012.
[7] IEC 60909-2. Short-circuit currents in three-phase A.C. systems. Part 2: Data of
ux = −Y 5 −1 · (Y 4 uphaS + Y 6 uphaR ) (C.8) electrical equipment for short-circuit current calculations, 2nd ed., 2008–2011.
[8] British Electricity International Modern Power Station Practice, EHV Transmis-
If Y5 is of large dimensions, its inversion can benefit from the sion, Permagon, Oxford, UK, 1991.
[9] R. Benato, A. Paolucci, EHV AC Undergrounding electrical power. Performance
partition techniques and parallel computing [3]. and planning Series: Power Systems, Springer, 2010, ISBN 978-1-84882-866-7
(translated also in Chinese).
[10] J.R. Carson, Wave propagation in overhead wires with ground return, Bell Syst.
References Tech. J. 5 (1926) 539–554.
[11] G. Kron, Tensor Analysis of Networks, John Wiley & Sons, Inc., New York, 1939.
[1] F. Calero, Mutual impedance in parallel lines – protective relaying and fault [12] C.L. Fortescue, Method of symmetrical co-ordinates applied to the solution of
location considerations, in: Proceedings of 4th International Conference on polyphase networks, AIEE Trans. 37 (Part II) (1918) 1027–1140.
Power System Protection and Automation, November, New Delhi, India, 2007, [13] F.M. Gatta, F. Iliceto, S. Lauria, New Formulae for the computation of zero
pp. I1–I30. sequence impedance of EHV overhead lines, L’Energia Elettrica 79 (2002) (in
[2] P.G. McLaren, I. Fernando, H. Liu, E. Dirks, G.W. Swift, C. Steele, Enhanced double Italian).
circuit line protection, IEEE Trans. Power Deliv. 12 (July (1)) (1997) 1100–1108. [14] E. Clarke, Circuit analysis of a-c power systems Symmetrical and Related Com-
[3] R. Benato, S. Dambone Sessa, F. Guglielmi, Determination of steady-state and ponents, 1, J. Wiley & Sons, 1950.
faulty regimes of overhead lines by means of multiconductor cell analysis [15] IEC 60909-0. Short-circuit currents in three-phase A.C. systems. Part 0: Calcu-
(MCA), Energies 5 (2012) 2771–2793, http://dx.doi.org/10.3390/en5082771. lation of currents, 1st ed., 2001–2007.