Also, the authors have described the various applications in which negative-sequence polarized ground relays may be utilized to advantage and have compared their characteristics with relays using other forms of polarization, which makes this paper a very valuable source of information when the engineer is faced with an unusual situation. This discussion suggests such a situation and the solution of its problems. The problem was caused by pole top fires on an ungrounded 55-kv delta-connected transmission network. The faults were generally arcing single line-to-ground faults that were very difficult to locate even by extensive switching. The ensuing outages frequently covered large areas that were disproportionate to the magnitude of the fault. It was believed that a better way of isolating faults than by manual switching could be found. Previous studies had shown that the cost of grounding the system properly and installing conventional zero-sequence current polarized directional overcurrent ground relays was prohibitive. The system involved was composed of nine interconnected major stations and approximately 325 miles of transmission lines, the longest of which was approximately 40 miles and the shortest approximately 7 miles long. An investigation and inventory of the stations in the system involved, disclosed that there were two stations near one end of the system with transformers whose 55-kv windings were connected in ungrounded Y and the neutrals were accessible, and that all but one of the other
stations were equipped with two potential transformers connected in open delta. Space limitations precluded the installations of a third potential transformer so that potential polarization could not be achieved. It appeared that only negative-sequence polarizing could be accomplished economically. From calculations, it was determined that overcurrent relay co-ordination could be obtained using zero-sequence current and that for negative-sequence polarizing, at two of the stations, the negative-sequence voltage would be low enough under some conditions to give marginal operation. A review of the combinations of conditions under which operation would be questionable revealed that they were far less frequent than those where operation would be ensured. On this basis, it was decided to install negative-sequence polarized ground relays at all points where zerosequence current was not available as a polarizing source. Although wrong relay operations occasionally could be expected at the two points mentioned, correct operations on the rest of the system would greatly improve the over-all operation of the system. Since the first installation, a small grounding bank has been installed at the far end of the system from the original wye. connected banks, not for polaizing, but for additional zero-sequence current for better relay co-ordination near the end of the system. Twenty-two negative-sequence polarized overcurrent ground relays have been installed to date, and in the past 4 years, 20 operations have been logged, with no improper operations. One of these occurred at a point where marginal opera-
tion was expected. The assumption of the calculated risk of wrong operations has been fully justified and the improvement in over-all system operation has been truly
gratifying.
W. A. Elmore and J. L. Blackburn: We appreciate the interesting discussions. They contribute materially to the paper by adding practical experiences and by lending emphasis that negative sequence is another valuable tool for relay engineers. In comment on the last paragraph of Mr. Hendrickson's discussion, the susceptibility of the relay to false tripping is extremely low and we are not aware of any operation problems in this connection. Several factors are involved to assure a high security to these relays: (1) the negative-sequence directional unit is used in the sense of a fault detector in conjunction with fault measuring or discrimination units; (2) the relays are designed to have negligible response to offset transients; (3) 3-phase loss of potential or 3-phase short-circuiting of the current transformers does not provide negative-sequence quantities; and (4) lightning and switching surges generally are not significant producers of negative sequence. Unequal loss of potential or current and unequal pole closing can produce negative sequence and could cause directional unit operation. In these cases the fault measuring or discriminating units and these inherent, or built-in, time delays would provide security at least until the second contingency of a fault occurred. We are not aware of any field problems in this area with the present relays.
fo
V IU*
tage
In
I1
11
I T . * | ransmissionl
i rc u i t s
J. L. BLACKBURN
MEMBER AIEE
the subject a little mysterious, particularly from a practical standpoint. This paper will review the fundamentals and study the positive-, negative-, and zero-sequence induction that can exist between circuits. Of particular interest is the negative-se-
with calculation examples for determining the positive-, negative-, and zero-sequence voltage induced in transmission circuits
paralleled either on the same tower or on adjacent towers. For nontransposed lines the induced voltages are a function of all
and negative-sequence mutual impedance
three sequence currents. The positive-
the zero-sequence self-impedance of the lines. The conductor configuration should be symmetrical around a vertical axis to
paper.'
Negative Sequence
is less than 10% of the self-impedance of the lines while the zero-sequence mutual
.U
tower or on adjacent towers in the same must be considered during -~~~right-of-way Paper 62-1073, recommended by the AIEE Relays fault calculations and in some cases in the Operations Department for presentation at the design of the protective relay system. AIEK Summer General Meeting, Denver, Colo, June 17-22, 106f2. Manuscript submitted March The fundamentals of mutual coupling, or 12, 1962; made available for prnting April6, electromagn.etic induction, have been L. BACKBRN i wih th WesingouseElecric stated in textbooks for many years. Corporaton, Newak, N. J.However, many enginleers appear to find
Electromagnetic induction between paralleled circuits exists because the flux established by the current in one line cuts the conductors of the adjacent or pa.rallel line. Consider two 3-phase parallel lines either on the same tower or on adjacent towers, as shown in Fig. 1, with only positive- and negative-sequence currents flowing. The net flux available from positive- or negative-sequence cur-
FEBRUARY 1963
921
b
a
'
Xb'
777777
7777777777777. ,77777
currents if it were physically possible to space each of the three conductors of the parallel circuit equidistant from each of the three conductors of the line itself. In other words, if a, b, c are the conductors of one line and a'b'c' the conductors of the other line, then the spacing d between the two circuits must be such that daa'=dab'= dac' and dba'=dbb'=dbc' and dcal=d,a,dec^ With this, the induced positiveand negative-sequence voltages are both zero. A study of this requirement will show that it is physically impossible to arrange the conductors to meet the requirements. The nearest approach would be to have each circuit exist as a triaxial cable. Thus, with practical separations of the conductors and circuits there will be induced positive- and negative-sequence voltages related to the degree of dissymmetry between the circuits. Practically, the positive- and negative-sequence induction between the circuits is very small and the mutual impedances are less than 10% of the self-impedance of the line. Usually they do not exceed 3 to 7%. These values are for nontransposed lines. Transpositions can reduce the values by 50% or more.
time phase. Actually, there would be no induced voltage in the parallel circuit from the positive- and negative-sequence
rent in either circuit to cut the adjacent circuit will be minimized because the currents are equal in magnitude and symmetrically spaced 120 degrees apart in
reference 3. Zero-sequence calculations are based on numerous assumptions including transposed lines and it is remarkable how close the calculated fault currents agree with values even with
line. The methods of calculating these self- and mutual zero-sequence impedances is well documented for transposed circuits and is tabulated for various circuits with and without ground wires in
quently, the mutual impedance ZOM between parallel lines can be as high as 50 to 70% of the self-impedance of the
and is cut by the flux established by current flowing in the line. Conse-
(_
777777777777777777
O I-2r'1O
'
the
untransposed lines.
Induction
Fundamentals of Mutual
Induction-Zero Sequence
pedances are complex quantities and except for the transposed zero-sequence equation (equation 57), their values may have angles anywhere between 0 and 360 degrees. This is because the induced voltages result from three 120degree electrically spaced currents and their relative physical proximity to the other three conductors in the adjacent circuits. Examining the general equations 18, 33, and 52, it is observed that all of the distances in the real term are involved in the square root denominator of the imaginary term. For the positive and negative sequence voltage equations 18 and 33, the real term of I, induced component of V2' is equal to but of opposite sign to the real term of the I2 induced component of V1'. Correspondingly, the real term of I2 for V2' is equal but of opposite sign to the real term of It for V1'. The imaginary term of I1 for V2' is equal to the imaginary term of I2 for V1', and correspondingly, the imaginary term 12 for V2' equals the A1 imaginary term for VI'. The zero-sequence components of the two voltages V1' and V2' are equal except for the opposite sign of the real component. In examining zero-sequence equation 52 it is observed that II and 12 components are equal except that the real functions have opposite signs. The zerosequence current factor is the symmetrical
1I = 0, and equation 52 reduces to equation 57 when I,=12= 0. The induced voltages or mutual im-
Zero-sequence current is equal and in-TalI phase in each of the three conductors and returns via the earth and/or the ~~ ground wires. This means that the three Apeno conductors of each circuit can be replaced by an equivalent single conductor, I as shown in Fig. 2. With the return II III mathematically averaging 2,800 feet IV
earth resistivity, it can be seen that the parallel circuit iS in the magnetic field
from the conductors for 100 meter-ohms v
blI
Induced Voltage
Linles
.....Untransposed
Equation Nos.
.
.
.
......... 1-18
1926
.......39-47
vi
viZr
zero sequence
euneTasoe.35
.....Untransposed .............48-52
.....Untransposed
922
FEBRUARY 1963
component formula for the mutual impedance as explained earlier in connection with equation 57 for transposed lines.
TYPIcAL EXAMPLES USING THE FORMULAS Two examples of the calculations of the induced voltages on parallel lines are given in Appendixes VIII and IX. The
duced voltage is found to be 51% of ZO. If these lines are not transposed most of the positive- and negative-sequence induced voltage is caused by the zerosequence current. With the assumptions of Io only and I,=12=lo the mutual impedances are 9. 1% and 7.73% of X1 respectively. These lines are unusually
spacings would reduce these values. Grouping the lines a-b-c-a'-b'-c' instead conductors are a', b', c'. To determine circuitline fiers wistforelatdble onpasinglbetower with relatively close spacings be-' the preferred a-b-c-c'-b'-a' changes the the induced voltage in Line 1, phases a, of tween the two circuits. The positive- and negative-sequence mutual reactance is induced voltages but not the percentages, b, c in Line 1 are changed to a', b', c' respectively and the phases used in Line -j 0 0168 ohm per mile, assuming that assuming Io only and 1 = 12 =Io. 2, corresponding to a, b, c, and the forboth lines are transposed. This is only EFFECT OF GROUND WIRES mulas used as developed. 2.34% of the self-reactance of either line. If the two sets of conductors are symThe presence of ground wires tends to By contrast, the zero-sequence mutual impedance with the lines transposed is 0.286 reduce the induced voltages. This is metrical around a vertical axis, then the +j 1.78 ohms per mile which is 63.8% of studied in connection with the induced induced voltages in the two lines will be the zero-sequence self-impedance, with negative-sequence voltage in Appendix V. equal. This is shown in part B of ApThe I1, 12, and Io expressions of equation pendix X. For either induction in Line the physical spacings as given. If the lines are not transposed, the posi- 47 are the same as the corresponding 1 or in Line 2, the spacings to the currenttive, negative-, and zero-sequence in- ones of equation 18, Appendix I, except inducing conductors in the adjacent line duced voltages are complex. In order to that equation 47 has additional factors for are the same. Where the conductors are not symcompare these, it was first assumed that the ground wire return currents. These ground wire currents, Ix and ly, are metrical around the vertical axis, as in the zero-sequence current predominates, This was accomplished by assuming 10 assumed to flow in a ditection opposite to Part A of Appendix X, the voltages induced in Line 2 from Line 1 currents are only with I2= 1 = 0. The second assump- Ia, I and I.. In the example of Appendix VIII, two different from the voltages induced in Line tion was to let I1=12= Io. With lo only, the induced positive- and negative- ground wires X and Y from left to right, 1 from Line 2 currents. A different voltsequence voltages are almost 1/2, and 24 feet apart and 13 above the middle age exists to cause circulation of currents with 1,=12=1o, the voltages are approxi- conductors b and b' were added, and the between the lines. For the nontransmately double that of the transposed case. induced negative-sequence voltage cal- posed horziontal lines of the Appendix X Increasing the spacings between these culated per -equation 47. Values for Ix example, the positive-sequence voltage and I are very difficult to calculate for difference is a function of I, and J0, and the two circuits reduces the mutual voltage. For example, if the spacing is doubled the untransposed circuits, but a good negative-sequence voltage difference of Z2m =j 0.0054 ohm per mile for transposed approximation can be obtained by assum- I2 and lo. The I, and I2 current factors lines and V2'= (-0.0047-j 0.00261) ing the circuits transposed, in whicb case are very small so that the effect essentially 1I-j 0.00542+(-0.00564+j 0.00326)Ie Ix and Il are a fraction of theIa current is a function of Io which exists during flowing. In the example 61.5% of the ground faults. The zero-sequence voltfor nontransposed lines. It is interesting to note that if the right- current will flow in the two ground wires age difference is a function of only I, and hand set of conductors are a', b', c' from assuming the ground wires are the same 12, so that a zero-sequence circulating top to bottom, instead of c', b', a', the type of conductors as the phase conduc- current can exist on phase faults, such as positive- and negative-sequence induced tors. If the right-hand, or Y, wire is an ab or ca fault, where I, and 12 are voltages are different but their total omitted, then 43% of the current returns not 180 degrees apart. mutual impedance values with either 1o in the X ground wire. These were deterGenerally, this unbalance effect has only or I1=12=Io are the same or 1.25% mined from reference 3. Table II not caused trouble but a case is known and 4.4% of X, respectively. The trans- summarizes several calculations based on where failure to trip occurred for an internal phase fault on a line using pilot posed values increase slightly from 2.34% equation 47, assuming I1121I0. wire relays having a very sensitive zeroto 2.5% of XI. The zero-sequence value INDUCED VOLTAGE UNBALANCES N THE sequence setting. The circulating zeroremained unchanged. The a, b, c-c' PARALLELED CIRCUITS sequence current provided a blocking b', a' top-to-bottom arrangement of the The formulas in the Appendixes give indication to the relays overcoming the conductors should be used since there are other reasons why a'b', c' top-to-bottom the induced voltages in the a', b', c' con- positive sequence. The solution is to conductor spacing is not desirable. These will be developed later in connection with Appendix X. Table II
The second example in Appendix LX
___________________________________
close together
and, again,
increased
ductors resulting from currents flowing in the a, b, c conductors. The formulas also can be used to calculate the induced voltages in the a, b, c set resulting from currents flowing in a'b'c' by interchanging the conductor labels a and a', b and b', c and c'. Calculations for a typical example are given in Appendix X. For calculating the induced voltage in Line 2, Line 1 conductors are a, b, c and Line 2
towers. With the lines transposed, the voltages are reduced. The positive- and negative-sequence mutual impedanlces are
1.09%o of X1 and the zero-sequence in-
Per Cenat of Xl
Per Cent
of
Xl
IX-IU =1/2IO 0053-1 0.0008. 75%. 0033+10.0032 . .....0. ....-0. ....0. 64% Iz=0.431o ..-0. l-0.0147-J 0.0111. 56%.... -0.0023+j0.0183 .....2. . ....2.56%0 IV/ - 0. 1851I0
Ix =IVi
1/210 does not aPPear PraCtiCal bUt iS included fOr comparative purposes.
FEBRUARY 1963
923
It should be noted that in the example of Appendix IX the voltages induced in the left line from currents in the right line will be different from those induced in the right line from the left line currents in all cases because the separations of the indivudal conductors in the two circuits are not the same. This would be true for parallel lines of different configurations.
abc-c'b'a'.
desensitize the relay or, preferably, to space symmetrically the two sets of conductors around the vertical axis. This is why it was previously stated that the preferred arrangement of the conductors in Appendix VIII is abc and c'b'a' from top to bottom and in Appendix IX,
Vb'=IaZab'+IbZbb'+IbZcb'
(2)
Correspondingly,
And the mutual voltage induced in conductor c', resulting from current flowing in conductors a, b, c, is
dba'+a2
(16)
Vc"=IaZac'+IbZbc'+IcZCC
in conductors a', b', c' from currents Ia, ba,h in conductors a, b, c is v,'=!( V'+a'Vb'+aVC') 3 1
=
(Zca'+a"Zcb'+aZcc')=
(Logdca'+alLogd,'+aLogdecc) (17)
0.2794
Substituting equations 15, 16, and 17 in equation 8, usng a- -0.5+j0.866 and as- - 0.5-i 0.868, and reducing and combining results in
Log dacdbbdca'
[Ia(Zaa'+aZab'+aZact)+
dab'dba'dec'
Ib(Zba,+aZbb'+aZbc)+ Ic(Zca'+a2Zcb'+aZcc')l
(4)
I + + Log -/Lgdab'dac'dba'dbb'dca'dcc'
Ia=I1+I2+10
Conclusions
The fundamentals of electromagnetic induction have been reviewed and the general formulas for the calculations of the positive-, negative-, and zero-sequence induced voltages in parallel transmission lines have been developed and applied to typical examples for transposed and
Substituting equations 5, 6, and 7 in equation 4 and reducing, gives 1/dab'dac'dbb'dbc'dcb drc '
Ib=a'I1+ah2+IO Ic = al +a'12+Io
(5) (6)
(7)
dabdca'dbc'dc'dcb']+
daa'dbadca'
,,,
1 (18)
dab'dbc'dca' daadbb'dec I
ages are complex and they are a function of all three sequence currents. The positive- and negative-sequence mutual impedances are less than 10% of the selfimpedance of the line and thus are negligible; however, the zero-sequence factor can be as high as 50 to 70% of the zero-sequence self-impedance and must be considered in fault studies and relay applicathe two linesthe are about a to symmetrical lms ar symmtrsc1 abut a
V2'=-Ii[(Zaa'+a2Zab'+aZa2')+ 3 a(Zba'+a2Zbb'+aZbe')+
as shown in the
3I0t(Z441+a'Z.t+aZa')+
following configurations:
Co
O
abo
(8)
(Zca`+a'Zcb'+aZcc')J
vertical axis, the voltages induced in the two parallel lines are equal. Nonsymmetry produces different voltages which can cause relaying problems. It is hoped that this study has contributed to a better understanding of voltage induction in transmission circuits.
Carson's formula' for mutual impedance With these transpositions, the mutual two paral3el a between with in conductors voltages induced a,in b, c' commo earh reurn, conductors t 60cycls, "resulting from currents Ia, lj,, c concommon eth return, at 0 cycles, i ductors a, b, c are as follows and similar Zaa=00954+j 0.2794 Log , ohms/mile to those of equations 1, 2, and 3: daa' 1
tions.Wihteetasoiin,temul
1st Section
cO
0,'
Oc
2nd Section
bor'
Oa
bIa
(9)
da
DI ohms/mile
V'3-(IaZaa'+IbZa'+IcZca')+ 1
3
1
UaZ-'+IbZcb+IcZab')+
3 (IaZcc'+IbZac'+IcZbc')
(19)
(11)3
~
1 v'= (IaZab'+IbZbb'+IcZcb')+
3
1
The conductors of one line are a, b, c, and of the other line a', b', c'. The disZaa'=0.0954+j 0.2794 Log D,tance d between the various conductors is j.74Lgda l) expressed by d,,' dab', dac' etc.; the j3.74Lgd, impedances by Zaa', Zac'. etc. a'Z,b'-0.09M4a'+j0.2794a' Log D,For any conductor arrangements, thej .74 Lgdb(1) .74 o a'(3 mutual voltage induced in conductor a' resulting from current flowing in conaZac' = 0.0954a+j 0.2794a Log Dcductors a, b, and c is j 0.2794a Log d,c' (14)
1- (IaZbc'+IbZcc'+IcZac')+
tactIZa+co)(0 1)-(hZa+tZa+Za)(0
1 Vc-(aZc+ZcIZc+ , IZc+bb'IZc) 1
- (IaZba'+IbZca'+IcZaa')+
Correspondingly, the mutual voltage induced in conductor 5', resulting from current flowing on conductors a, b, c, is
(Zaa'+ a'Zab'+aZac') =-j 0.2794 (Log d,,'+a' Log dab'+a Log d,') (15)
(IaZca'+IbZ,a/+IcZbb') (21)
FEBRUARY 1963
924
VI' 3 [Ia(Zaa'+aZab'+a2Zac')+
bZa'ab'aZc' IV(ZbI+aZbbaZbC)
equation 28, and reducing gives 1 VI1 - I [(Zaa'+aZab'+ a2Zac')+
Ie(Zca'+aZe&'+a2Zcc')1
(28)
in conductors a', b', c' resulting from currents Ia, Ib, I, is 1 (34) l' -( Va'+a Vb'+aIVc')
V2'=1 9 [Ia(Zaa'+Zub'+Zcc')+
Ii(Zba'+Zcb'+Zac')+ Ic(ZCa'+Zab'+Zbbc') +
1 - [a'Ia(Zab'+Zbc'+Zca')+ 9 a'Ib(Zaa +Zbb +Zcc')+
a
VI' -a(Zaa"+Zbb'+Zcc)+ 9
=
Ib(Za'+Zcb'+Zac')+ Ic(Zcea'+Zab'+Zbc')I +
a'(Zba'+aZbb'+a2Zbc')+
a(Za'+aZcb'+aZcc ')] +
+
+Ic(Zc+'+Zac'+Zba') + a'(Z,a'+aZcb'+aZcc')JI+
1 9
[a1a(Zac'+Zsa'+Zcb')+
aIc(Zaa'+Zbb'+Zcc')
1
a1b(Zbc'+Zca+Zasb)+
(22) 7in Substtutin equaions5, 6,and 7 in3 Substituting equations 5, 6, and reducing, equation 22 and combining and reducing, the positive- and zero-sequence current factors all cancel out since 1+a+a2 =0. Therefore,
3O1(Zaa'+aZab'+a2Zac')+
V2
1 - I2[(Zaa'+Z'+Zcc')+ 3
Applying Carson's formulas' in a manner similar to that shown in equations 9 through 17, new values for the impedance combinations are obtained as follows:
By substituting equations 5, 6, and 7 in equation 35, and combining and reducing, the negative- and zero-sequence current factors all cancel out since 1+a+a$ -O.
Thus,
1 VI -- IIE(Zaa'+Zbb'+Zee')+ 3
a(+Zc51+Za 'Zc')
V2'=- Id(Zaa'+a'Zab'+aZac')+
3
(3)
It will be noted that equation 23 is the same as the 12 component of equation 8. Another form of equation 23 is
a(Zab`+ZbC'+Zea')+
combining
one
a'(Z '+Zsa'+Z')]
(36)
(32)
W=0.09311, 0.866Log
j
a1+(Zea'+a2Zcbt)+aZcc'j (24 (24) a(Zc01+a9Zcs'+aZ~c')] Substituting equations 15. 16, and 17 in equation 24, combining and reducing
(
gives
0.2794 F
3
Substituting equations 30, 31, and 32 in equation 29 and reducing and combiniing, remembering that a--0.5+j O.86W and a= 0.5 j0.866, gives
Lo09dbdc
daadbb'dcb Vdab'dac'dbadbc'dca'dcb'J
r
Finally,
1
daa'dbb'dcc'
ZIM
31VI0da68dLog,,,0.0931 8
'bdac'da'
Finally,
Z
0.2794 23' =
=-
Ldab
'
dac'daa'dcb' _
dadbc5'd5a'
1
j Log (38) /dab'dac'dba'dbc'dca'dcb'l Log (38 jL o dac'd,b'dca' where ZiM is the positive-sequence mutual 0.2794 , impedance in ohms/phase/mile at 60 + Vdab 'dac'dba 'dbbdcadcc cycles.
d25
Logjd
\1dab'dacdba'dbc'dcaLdcbog dac'doc'dcc'
(26)
dbd'
positive-sequence
33
V'dab'dac'dba'dbc'dca'dcb'
Appendix V. Negative-
in2V')2
conductors are 4,v', dab~', d4ss, etc., so that derivations apply to varying configurations.
shown:
z(gl)
a0
bo05'
c0
a
FEBRUARY 1963
925
0.093110
gd
dac'd&c'dcc'
'
IzZxaI'Zva' (39)
Vb' =IaZab'+IbZbO'+IcZcb'-
j Log
daa'dba 'dcs
dca'dcb'dcc'
j0L[gdLoagd5a'da
j Log
dza'
0.0931Iy
i Log
in conductors a', b', c' from currents Ia, Ib, Ic in conductors a, b, c and -I. and -II in the ground wires x and y is
The induced
negative-sequence voltage
Zyl(41)
dd'd'b'dc'J
v2=1(vn +asvb/+avc/) 3
1
3
0.2862+j 0.8382
Log
Appendix
da'
volts/mile at 60 cycles (52)
[Ia(Zaa'+aZab'+aZac')+
Ib(Zba+a'hZabb+aZbct)+
Ix(ZZa'+a2Zxb'+aZxc')-
Ic(Zca'+a2ZctO'+aZcc')-
Il/(ZYa'+a2Zyb'+aZyc')] (42)
(48)
Indix Induced
Vii.
Volt
Zero-Sequence
Foro
Two
The first three terms of equation 42 are Substituting equations 1, 2, and 3, gives, Assume that the lines are transposed as the same as equation 4 and thus reduce outlined in Appendix II. The voltages to equation 18. The last two terms with 1induced in conductors a', b', c' resulting 703 [Ia(Zaa'+Zab'+Zac')+ Ix and Iy can be reduced applying Carson's from current flowing in conductors a, b, c formula similar to that in equations 9-11. fare given in equations 19, 20, and 21
(Zza'+a'Zzb'+aZzc') -j 0.2794
=
lb(Zba+Zbb'+Zbc')+
(Zya'a2Zvs'+aZyc') = -j 0.2794
Further substitution of a-0.5+j 0.866 and a= -0.5-i 0.866 in equation 43 and 44 gives
Ic(Zca'+Zcb'+Zcc')J .
(49)
Vs' II[(Zaa0'+Zab'+Zac')+
1 3
(53)
a2(Zba'+Zbb'+Zbc')+
1
a(Zca'+Zcb'+Zcc')]+
3 2aaa+ab+ac)+
Vo"'
Ia(ZaG'+Z55'+Zcc')+
Ib(Zba
Ic(Zca'+Z00'+Z&c')]+
+ZcbD+Zac')+
a(Zaa'+Zbb'+Zbc')+
a2(Zca'+Zcb'+Zcc')] +
(Zba +Zbb +zbc )+ Z+ c)]
1[a(Zab'+Zbc'+Zca')+
0.2794(+0.866 Log
j
-dra
1Io[(Zaa' +Zab1+Zac')+
(45)
\Ldzbgdzzc)
d\
b
((aZca, Z+b'
(50)
06J(46)
(46
By substituting Carson's for mutual impedance, as given formula2 in equations 10, 11, etc., and reducing and collecting 9, terms, equation 50 becomes
Thus,
(54) By substituting equations 5, 6, and 7 in equation 54, and combining and reducing, the positive- and negative-sequence current factors all cancel out since
Ic(Zaa'+Zab'+Zcc')1
1+a+as=0.
t 1
V0 3 0Za+b+ZcZa+
Applying Carson's
V21
dab'dac'dba'dbb'deczdcc
daatdbctdobt
dba'dbb'dbC'dca'dtdcrdc'+
Log
d001d01,'d001+
formulas' same with 3manner equations as 9, shown 10, andin11,connection the impedance
expression of equation 55 becomes
0.0931I2F0.866 Lgdac'4ba'dcbt
.6(.29)Lg'btbto
jLog /a
dAaa'db5'deCC' +
c tde dctc dbt
Laotc
VOM
1095+.24Lo
926
FEBRUARY 1963
j 0.2794 Log D, -j 0.2794 Log dbb'+ 0.0954+j 0.2794 Log D. i 0.2794 Log d,A'] and so forth for the remaining 6 terms] (56)
Log
Zo
91
~~~~~~~~~~~~.03
-;/daa'dab'dac'dba'dbl'dbc'dca'dcb'dec'
ohms/mile at 60 cycles (57)
________________De
0.0168
~~X 072
Z.
This equation is derived similarily as the GMR-GMD concepts in symmetrical components, see references 2 and 3.
I, = I2 = Io:
V2 =(-0.0234-j 0.0213)12
4.4% of XI
f Z2M = -j 0.0168 ohms/mile 64.3% of (26) /80.9 degrees, Zo 1.811. the Positive-, Negative-, and Zero- Zom 0.286+j 1.78 1.80 /80.9 degrees If the right-hand set of conductors are Sequence Induced Voltages in ZOM=0.286+i1.78-1.80 /80.9 degreeS a', b', C' from top to bottom instead of ohms/mile (57) a' as previously studied, the results Two Parallel Lines on the Same c', b', _ different and are aS fOlloWS: ~~~~~~~~~~~~~~~~~~~are
=
Appendix
VINI.
Calculations d
(38)
VO'(0.286+j 1.789)Io
Tower
The configuration and spacings are shown in the sketch. For these lines
dac' = dCa'18 feet
As indicated above ZiM and Z2M are 2.34% of the positive- and negativesequence reactance (XI =X2) of the line
compared with ZOm which is 63.8% of zero-sequence impedance ZO.
the
dab'= dba'= dbc'= deb'= 22.5 feet daa'=dbb'd,cc=24 feet Ra 0.259, Xa = 0.435, Xd for 3 conductors
0.285
(-0.0078+j 0.0045)Io
(18)
0.4 18'
8'
Go..8~C'
Log
24 X22.5X22.5
8'
I04'8
.
By previous calculationss, the constants of the line, assuming that it is transposed, are
24X24X24 /22.5X18X22.5X22.5X18X22.5
F
18X22.5X24 g 22.5X24 X22.5 ] 24 X22.5X 18 volts/mile (18)
+0.0931o 0.866
V2' = (0.0068+j
Vo'=(0.286+j
Appendix
V/22.5X18X24X22.5X22.5X24
and zero-sequence voltages are calculated from equations 33 and 52. All induced voltages in volts/mile are summarized.
V,' = -j 0168II+(0.0156-j O.009)I2+
Zorn=0.286+j 1.78=1.80 /80.9 degree ohms/mile, mutual Paralleled Zo of both circuits =0.415+
oo~ ~
Xomn6 -=41
X
t-38 -=.8 X,
the
Positive-, Negative
IX.
Calculations of
and
Lines
The configuration and spacings are shown in Fig. 3. For the lines inl Fig. 3
a'
Fig. 3. ConFiguration and spacings for positive-, negative-, end zero-sequence induced voltage in two parallel lines on separate towers
FEBRUARTY 1963
.927
daa'- 105 dba'=92 dca'=79 dab'- 83 dbb'-70 dcb'=57 dac'- 61 dbc'-48 dec'=35
J Lo L
105X70X35
1+
Vo'=(0.286+j 1.29)Io=1.321o
/77.5 degrees, 48.5% of Z,
Z1 -0.306+j 0.784 =0.84 /68.7 degrees ohms/mile Zo-0.592+j 2.675=2.72 /77.5 degrees
ohms/mile
volts/mile (18) Similarly, using equations 33 and 52, the induced positive- and zero-sequence voltages can be calculated. All three voltages are summarized:
j Log fe a83X61X70X48X57X35
105L3X79X57
1
Zi
3.24
V2'=(0.0082+j 0.0018)I1+
(0.041-j
0.0575)Io
Line l b c
a'
b'
'-0.093112
iLog-
(-0.041-i 0.0575)1o (18) Vo'=(-0.0236-j 0.0356)Ih+ (0.0236-j 0.0356)I2+ 0 (0.286+j 1.36)Io (52) For comparisons: lo only I1 =12=0 VI'=0.0711o /-54. 5 degrees, 9.1 % of XI
V2 '=0.07110 /234.5 degrees, 9.1% of XI VO'= 1.391o /78.2 degrees, 51% of ZO For 11=I2=1o
(0.00057+j 0.0085)I2+
da4'=61
0 13' 0 13' 0
dbaj'=48 dca'-35
d"' =61
35'~
c'
0 13' 0 13 0
dab' = 74
dac' = 87
d,b,' =48
dcct= 61
dbeI 74
VI'=(-0.00157-j 0.00593)I,+
V/83X61 X92X48X79X57J
105X70X35
V,'=(-0.0322-] 0.0472)12=0.0571I2
/235.7 degrees,
0.0085+
074100=1.09% of positive0.784
7.3% of Xi
at
(0.286+j 1.41)1,
Vo'=(0.286+j 1.29)Io=1.321o
/77.5 degrees, 48.5% of Zo
sequence reactance Substitution in equations 38, and 57 gives the positive- and zero-sequence mutual impedances. A summary of these values for the three sequences is as follows: Zjm - -0.00057+j .0085=0.0085 /93.8 degrees ohms/mile (38)
If the right-hand set of conductors are a', b', c' from left to right instead of c' b', a' as previously studied, the results are different. For the nontransposed lines V1'-(-0.0025-j 0.00795)I1+ (0.00705-j 0.00474)I2+
b'
daa'"61
d,,'= 48
4a'=74
dbb'=61
dca'=87
dc '-61
de'=74
ZoM=0.286+j
1.36 = 1.39
V1'=(0.00157-j 0.00593)hi+
(0.00555-jO.00311)I,+
=
Zim and ZIM are 1.09% of the positiveand negative-sequence reactances (X1-X2) of the line compared with ZOM which is 63.8% of the zero-sequence impedance (ZO).
(-0.0236]j 0.0356)Ii+
(0.0236-j 0.0356)I2+
(0.0262-j 0.0392)I+
(0.286+j 1.36)Io
9.1% ofX1
V83X61 X92X70X79X35J+
X92 X57 og61 83X48X79
~~105X48X57
-+
.O31I[O 66
VKs
(-0.047+] 0.0815)1,
928
FEBRUARY 1963
Line 1 Line 2 a' b' c' c b a 13' 013' 0 35' 0 13' 0 13' 0 O
7 d dab'=74 dbb'=61 dcb'=48 = dac' 61 dbc' = 48 dc0'= 35 Voltage induced in either Line 1 from curent'm in 2, 2,or m Line Lne2 fom current urrnt current in Line or in 2 from
daa'=87 dba'=74
dccz'=61
=jO.1631i.
References
1. NBOATIVR-StQuBcNcB DiRBcTrIONAL
RELAYING, W. A. Elmore, J. L. Blackbur GtowND AIEX Transadioxs, see pages 913-21 of this issue.
2. SYXMBTRIcAL Wagner, R. D.
F. (book), C. CompoNENqTa Book Evans. McGraw-Hill Company, Inc., New York, N. Y., 1933.
S. SPBED ZBRO SEQUINCE IMPEDANCE CALCULATlom Font TRANSMuSSsON LINEsS, J. L. Blackburn Eectrical World, vol. 163, March 1960, pp. 60-61.
V2 =(0.00592+j 0.00161)I +
rotection
or
wircuits
e .
wa7pa cl to rs
ASSOCIATE MEMBER AIEE
in some instances can even cause incorrect operation of the differential type of line
relaying such
as
JOHN BERDY
This section will briefly review this subject and in addition discuss the effects of the series capacitor protective gaps.
phase comparison.
Summary: An analysis of the protective problems introduced by series capacitors leads to the conclusion that conventional line protective schemes will, in most instances, have limited application on seriescompensated lines. A line relaying scheme is presented which has almost universal application on compensated lines. This scheme, designed specifically for lines with series capacitors, will meet the speed, selectivity, and sensitivity requirements of present-day systems.
pensation can provide improved stability voltage regulation, a de=imits, improved p o a onomic s, sied load divsion, oor ra more economic
a
utilization of the
extra-high-voltage sys-
I tors to high-voltage transmission lines has once again been receiving considerable attention during the past few years. For the most part, the renewed interest in this subject has been in the area of extra-highvoltage transmission. With a few exceptions, most of the extra-high-voltage systems being planned for construction in the United States and in Canada are being designed to use some degree of series com-
pensation.
The benefits of series capacitor compensation from a system design and operating point of view are well known to the system planning engineer. Series cornPaper 62-1093, recommended by the AIEE Relays Committee ad approved by the AIEE Technical Operations Department for presentation at the\ AIRE Summer General Meeting, Denver, Colo.,
June 17-22, 1962. Manusrpt submitted Dccember 11, 1961; made avallable for printing,
and operation is the protective relaying of the series-compensated system, for, in most instances, the protection of circuiuts with series capacitors is not as simple and straightforward as the protection of other circuit configurations. Depending on its location and on the degree of compensation, the series capacitor may introduce various relay problems which could result in increased fault-clearing times and/ or a loss of selectivity. It is the purpose of this paper to discuss the protective problems associated with lines having series capacitors and to discuss the applicability of standard protective schemes to such lines. In addition, this paper will present a relaying scheme which was specifically designed for the protection of lines with series capacitors and which meets the speed, selectivity, and sensitivity requirements of present-day systems.
Effect of Series Capacitors Protective ReLays\
tion of directional-type protective relays by affecting the phase position of the voltages and currents as seenbytherelays. When a series capacitor is located between We eiscpctri oae ewe a relay terminal and a fault, the relay currents or voltages may be reversed from the positions they usually assume for a fault on a system whose reactances are inductive. For instance, consider the equivalent system shown in Fig. 1. For discussion purposes, assume that the system voltages are m phase prior to the fault and during the initial stages of the fault and that the system angle is 60 degrees, inductive. Fig. 1 shows the vector diagramof the system voltages and currents on this inductive system for a 3-phase fault (F) near terminal A. The vector diagram also shows the voltage VA which would appear at the terminals of a relay
IFI
A-
f2
F
2
VA
on\
Ma . 92
.IF The adverse effects of series capacitors on protective relay operation have been
Schenectady, N. Y.Ithsbeshw
COmpany,
previously
I a
ensonta sre
described
Fig.
1.
V/ector dibgram
1F
for an
Inductive
929
FEBRUARY 1963