Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
(1) Padova University, Italy (2) AREVA T&D S.p.A., Italy (3) Cardiff University, UK (4) National Grid, UK
ABSTRACT
Computation of capacitive and inductive coupling between parallel circuits of power lines allows determination of
induced currents and voltages on the non-energised circuits due to the voltages and currents on the adjacent energised
circuit(s). Various computer tools are now available for these computations. However, there is still some uncertainty
with regards to accuracy of results and capabilities of each software. For example, some programs have limitations with
the maximum number of circuits that can be analysed. These programs first require the line parameters to be determined
and then implemented in circuit simulation of the lines including the effect of span length and number and line
terminations.
In this study, various approaches have been considered and compared namely: SimPowerSystem, EMTP, MicroTran
and newly developed routines to calculate the line parameters (R, L, C and Z) required for circuit studies. The
calculation principles for each method were examined and correlated to the error analysis.
Keywords: Induced currents and voltages, parallel power lines and electromagnetic coupling.
303
impedance.
The first terms of equations (2) and (3) give the
jw2104(InDe\F&2
u Rearth+I w2 O nJ51[ J
=R
catenary curve of each span is evaluated. Subsequently, The capacitance matrix is then found by matrix
each span is discretized into short multiconductor inversion.
elements. In this way, each span is represented as a
series of "nominal Pi" blocks (cells) where it is possible
[HPT-1,
[
FF1
(
km
to apply the hypothesis that conductors are parallel to The procedure evaluates the R, L and C parameters for
the ground. The number of cells for each span can be each cell of all conductors [3, 4]. Cascade connections
adjusted easily. The catenary tool can be very useful in of multiphase Pi-circuits can be easily used to model
particular evaluations, such as long lines with different each span and, consequently, the full length of the line
span lengths and tower heights. Once the tower type and without any limits in the number of phase conductors
the tension force have been defined, the actual sag is and in particular, for parallel untrasposed lines on the
calculated. same right of way.
The evaluation procedure of resistance R, inductance L The load is represented by an equivalent shunt
and capacitance C for each cell, is explained in the impendance in cascade with the last cell of the system
previous section. obtained. If thereare constant-power loads, an iterative
The internal, external and mutual impedance of each method would be necessary to evaluate the equivalent
conductor are obtained according to Carson's formulae current injection depending in turn on the nodal voltage.
with the following assumptions: the conductors are Forrour purposes, theecurrentnisuestimated withlreferene
parallel to each other and the earth is homogenous to the nominal vtae.
within a span. The correction terms AR and AX have The generation block is described by the nodal applied
been programmed in order to change easily the number voltage vector [B]. This complex vector is formed by
of terms taken into account. two parts, the first one defined by voltage (in magnitude
An alternative method has been introduced in the RoMo and phase angle) applied by generators, the second is
procedure, according to Carson-Clem's formulae [2], unknown.
reported below, in order to compare the accuracy of
both methods under different conditions . 1 2I__l
Carson - Clem's Formulae 2 2
These simplified expressions are the first term of series 3 Lines 3
of AR and and the first two term of the series of AX. .........
The internal and external reactance of a conductor is son ri n
given by: Ge Res) Re G
Load r
304
The induced electromotive forces and consequently special cases, measured values can be used instead,
currents at the sending and receiving-end are calculated overcoming this approximation.
by studying the equivalent nodal admittance matrix Y SimPowerSystem operates in the Simulink®
(2nx2n). environment [5], and it allows the power system
The network is described by the linear system: transient and steady state to be simulated.
[I]=[Y][E] (11) The line model adopted in the case study was the n-
where phase distributed parameter line model with lumped
[I]: the nodal injected currents losses. The model is based on one used by the
[E]: the nodal applied voltages Electromagnetic Transient Program (EMTP) [6]. The Pi
By partitioning the nodal admittance matrix [Y] into model in Simpower was not used because it uses only
"known" and "unknown" voltage nodes ("k" and "u" the symmetrical components.
suffixes) the system can be described by: Instead, in EMTP and MicroTran [7], the n-phase
kIk ] [Ykk
= Yku ]0Ek 0 ( l2 ) nominal Pi model was adopted in the case studies.
iu uk Yu,_Eu
Y
i
(~~~~~~12)
3 CASE STUDIES
This system yields the following expressions of the
unknown voltages [E.]: Basic case
[E.l I = k.u1] I,u]- k ][Ek ]} ( 13) The first case
circuit.
study is a single span of a 400 kV double
It is known that the vector Iuu is equal to 0, because the Span Length: 0.366 km,
[Y] matrix represents the whole passive system Geometrical tower data: L2 configuration.
including the loads. Hence, it is possible to obtain the Conductors: Zebra (phase conductors), Lynx (earth
following expression: wire).
Eu- K-.Yuu yuk Ek ( 14) The load at receiving end of 1St circuit is assumed to be
a simple resistive load (1750 MW ).
Then, assuming that the equivalent nodal admittance One circuit is energised while the second one and the
matrix [Y0] represents the whole network without any earth ground wire are earthed to ground with a 0.2 ohm
load, it is possible to evaluate the currents at the resistance (typical value for a substation earthing system
receiving-end port, with the following expression resistance) at the sending-end and receiving-end.
[I]=[Yo][E] ( 15 ) The evaluation of line parameters for all programs is
[T]e[Yr]ce]ure be easily appliedat
(15)
section to
shown in Table 1. The relative error in all comparisons
ection.is
can
Thaluateproedure
evaluate voltagecandburreneasilytapp
and current at anyat span.
any s very small. The smallest one is between MicroTran
and EMTP, as it was expected since the two programs
are developed by the same authors. The skin effect is
2.3 Aspects of calculations in EMTP, Simpower and
not taken into account in RoMo, but it affects at
MicroTran MicroTran
In these programs, the resistance and the inductance are
~~~~~~~~~~nominal
condt
system frequency (50 Hz) with
sistance.Tsncrease
an increase of
with respet oD
evaluated taking into account the ground resistivity and conductor resistance. This increase with respect to DC
frequency and introducing some correction terms .anc lu s, the ma.xImum erowh wr s
accordng to Carson's Caron's
accordinto formulae.and
formulae.effect is
C evaluations, the maximum error without skin
0.7%.
In all programs, the catenary curve is not evaluated. The
conductors are assumed parallel to the ground at an Table 1 RLC parameters comparison.
average height given by: MicroTran Relative Relative Relative
(2 Ymin + Ytwr) Vs Resistance Inductance Capacitance
Yave =
where
3 [in]1
(16)% RoMo
Error
1.02
Error
0.04
Error
0.02
Ymin is the height of conductor at mid span Simpower
EMTP
0.64
0.01
0.05
0.01
0.09
0.03
Ytwr is the height of conductor at tower.
So, it is necessary to evaluate with the other program In addition some practical configurations can affect the
the actual span or use an average one. above values in real power line, e.g. the presence of
The EMTP ( Electromagnetic Transient Program) and joints, the actual length of the conductor due the
SimpowerSystem programs allow the user to define the catenary curve. The increase of DC resistance can be up
network easily with a graphical interface. A specific 2% for these factors.
routine is provided for tower geometric definition and The steady-state evaluation with RoMo gives the results
RLC parameter evaluation. shown in Table 2. In this example the conductor are
All these programs can take into account the skin effect, overloaded in order to cause a larger value in the
the steel-reinforced aluminium cables (ACSR) are induced voltages and currents, making easier the
approximated as tubular conductors. However, for comparison between the two programs.
305
Table 2 RoMo output, straight conductor configuration. (G bus RoMo are shown in Table 4. The relative error between
name at sending end, GW earth ground wire) MicroTran and SimpowerSystem for the 14-phase
RoMo _ conductor is less than 9%.
v I
Conductor Magnitude Phase Angle Magnitude Phase Angle Table 4 Voltage and current at receiving-end with
[VI [degrees] [Al [degrees] RoMo procedure. (E bus name at receiving end)
08 230900.0 -0.1 2525.5 -0. _ RoMo Receiving end
G9 230880.0 119.9 2525.3 119.v
G10 230840.0 -120.1 2524.9 -120.1 Conductor Magnitude Phase Angle Magnitude Phase Angle
G11 10.7 -86 53.4 -8 IV] [degrees] [A] [degrees]
G12 3.0 26.8 14.9 26. E1 0.82 43.8 4.10 43.
G13 11.8 97.7 59.0 97. E2 1.21 39.6 6.04 39.
GWG14 | 18.1 -95.9 90.7 -95| E3 2.00 41.6 10.00 41.
E4 0.97 49.0 4.86 49.
E5 1.52 43.1 7.62 43.1
The relative error between voltage and current E6 2.55 43.7 12.74 43.
magnitude and phase angle respectively of each GWE7 0.45 69.4 2.23 69.
conductors are shown in Table 3. As can be seen, good E9 229230 117.5 2507.20 117.
agreement is obtained. E10 228130 -122.7 2495.10 -122.
Eli 22.78 -140.6 113.89 -140.
Table 3 Voltage and current at receiving end E12 10.88 -20.5 54.41 -20.
comparison, MicroTran vs. RoMo (relative error). E13 33.17 38.5 165.84 38.
GWE14 38.61 -143 193.02 -14
Receiving end
AV% Al%
Magnitude Phase angle O°Magnitude Phase angle 0 For this configuration, the induced voltage and current
0.0% 0.0% 0.0% 0.00/ values in the first line placed 100 m apart from the
0.0% 0.0% 0.0% 0.00/ second line (Ist and 2nd circuits) are very small. Table 5
-0.1% 0.2% -0.1% 0.20/ shows a good accordance between MicroTran and
-0.5% 1.1% -0.5% 1.1 RoMo results in this case too.
0.0% 0.1% 0.0% 0.1
-0.8% -0. 1% -0.8% -0. 1 Table 5 Voltage and current at receiving end comparison,
MicroTran vs. RoMo (relative error).
Extended case Correction Term DR and DX
Another case study is undertaken using two parallel Receiving end
overhead lines, each defined as a typical 400 kV L2 AV% Al%
Magnitude Phase ang. Magnitude Phase ang.
double circuit overhead line. The lines are 10 km long, 5.8% 2.6% 5.8% 2.6%
with no transposition. The axial distance between the 3.2% 0.6% 3.2% 0.6%
2.0% -0.5% 2.0% -0.5%
two double circuits is 100 m. 6.6% 1.9% 6.6% 1.9%
3.6% 0.8% 3.6% 0.8%
1st circuit 2.3% 0.0% 2.3% 0.0%
-8.9% 5.0% -8.9% 5.0%
_ _, jj __ 0.0% -0.1% -0.1% 2.7%/
2nd circuit 0.0% 0.0% -0.1% 0.1%
__| 0.0% 0.0% 0.1% 0.0%
: ~~~~~~~~~~ ~~~~~~-1.4%
0.1% -1.4% 0.1%
100 m 0.4% -10.2% 0.4% -10.2%
0.9% 0.8% 0.9% 0.80/
-3.4% -0.2% -3.4% -0.2%
0 3rd circuit
._..............................................................
rL In order to investigate the sensitivity of the method
lI4th4th circut circuit adopted in the evaluation of induced voltages and
Figure 2 Representation of the case study. currents due to changes of the value of the soil
resistivity, some computer simulations were performed.
Tower footing resistance:1.0 Q The comparison of results shows a relative error of 30%
Earth substation resistance: 0.2 Q between RoMo with 24 correction terms of AR and AX
Span length: 0.366 km, no. of spans: 27 and RoMo with Carson-Clem's formula in case of a low
Conductors and tower configuration: (as above in basic earth resistivity (20 Q m). Indeed, during the
example). sensitivity analysis the assumptions in Carson-Clem's
In this case, only the 3rd circuit is energised, while all formulae (ratio dmax and De less than 0.15) was not
the other three circuits are earthed at both end. fulfilled in this case. For the case of 20 Q m earth
The load at the receiving end of the 3rd circuit is resistivity, the ratio was 0.30, double the limit. In this
assumed to be a simple resistive load of 1750 MW, case the results are affected by a larger error, and the
The earth ground wires are earthed at the sending and first method is needed.
receiving ends through the substation earth resistance, In SimPowerSystem, it was easy to monitor all currents
and they are earthed at each tower along the spans. The that flows through the tower footing resistance. The
tower footing resistance is assumed to be 10 Q2 for all distribution of the earth current along the tower line is
towers. shown in Figure 2. The currents through the first 14
The results obtained with the developed procedure towers have a different direction than the others. In the
306
middle, the current is almost zero. This evaluation was The proposed procedure is flexible, easy to use and it
possible adopting MicroTran, SimpowerSystem and has shown speed and accuracy in the evaluation.
RoMo, and results with a good accordance were A comparison with well established software packages
obtained. for power systems studies has shown satisfactory
E W condutor: 2 agreement.
4 , The limitation of approximate formula is confirmed in
2- the extended case. Indeed, in order to investigate a wide
E
range of earth resistivity values and in case of wide
-2- ~~~~~~~~~~~~~~~~~~insCaronsserie
separation, edd
more terms in Carson's series are needed.
10 15 l llAlfurther step of this work will be the comparison
0 5 10 15 20 25 between simulation results and measurement data of
100 operating electrical power line.
C_
t -100-
/REFERENCES
307