PowerFactory 2024

Technical Reference

Overhead Line Models

ElmLne, TypLne, TypGeo, TypTow
F2024

POWER SYSTEM SOLUTIONS

MADE IN GERMANY
 Publisher:
DIgSILENT GmbH
Heinrich-Hertz-Straße 9
72810 Gomaringen / Germany
Tel.: +49 (0) 7072-9168-0
Fax: +49 (0) 7072-9168-88
info@digsilent.de

Please visit our homepage at:

https://www.digsilent.de

Copyright © 2024 DIgSILENT GmbH

All rights reserved. No part of this
publication may be reproduced or
distributed in any form without written
permission of DIgSILENT GmbH.

January 9, 2024
PowerFactory 2024
Revision 1
Contents

Contents

1 General Description 1

1.1 Mutual Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1.1 AC/DC Couplings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Frequency Dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.3 Lumped and Distributed Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.4 Line Shunt Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.4.1 Input Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Equivalent Circuits 5

2.1 Lumped Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.1.1 Single-Phase Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.1.2 Single-Phase Line with Neutral Conductor . . . . . . . . . . . . . . . . . . 7

2.1.3 Two-Phase Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.1.4 Two-Phase Line with Neutral Conductor . . . . . . . . . . . . . . . . . . . 9

2.1.5 Three-Phase Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.1.6 Three-Phase Line with Neutral Conductor . . . . . . . . . . . . . . . . . . 11

2.1.7 4-Wire Model Data Conversion . . . . . . . . . . . . . . . . . . . . . . . . 11

2.2 Distributed Parameter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.2.1 General Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3 Load Flow Analysis 19

3.1 Temperature Dependency of Line Resistance . . . . . . . . . . . . . . . . . . . . 19

3.2 AC Model Calculation Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.3 AC Model Calculation Parameters for Linear DC Load Flow . . . . . . . . . . . . 22

3.4 DC Model Calculation Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.5 Consideration of Line Shunt Compensation . . . . . . . . . . . . . . . . . . . . . 24

4 Short-Circuit 25

4.1 Zero sequence mutual impedance . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.2 Consideration of Line Shunt Compensation . . . . . . . . . . . . . . . . . . . . . 27

DIgSILENT PowerFactory 2024, Technical Reference i

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
Contents

5 Harmonics Analysis 28

5.1 Zero sequence mutual impedance . . . . . . . . . . . . . . . . . . . . . . . . . . 28

5.2 Consideration of Line Shunt Compensation . . . . . . . . . . . . . . . . . . . . . 28

6 RMS Simulation 29

6.1 Zero sequence mutual impedance . . . . . . . . . . . . . . . . . . . . . . . . . . 29

6.2 Consideration of Line Shunt Compensation . . . . . . . . . . . . . . . . . . . . . 29

7 EMT Simulation 30

7.1 Lumped Parameter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

7.2 Distributed Parameter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

7.2.1 Bergeron’s Method for Solution in the Time Domain . . . . . . . . . . . . 31

7.2.2 Constant Parameter Model . . . . . . . . . . . . . . . . . . . . . . . . . . 33

7.2.3 Frequency Dependent Parameter (Modal) Model . . . . . . . . . . . . . . 35

7.2.4 Frequency Dependent Phase-Domain Models . . . . . . . . . . . . . . . 37

7.3 Diagonalisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

7.4 Consideration of Line Shunt Compensation . . . . . . . . . . . . . . . . . . . . . 41

A Parameter Definitions 43

B References 46

List of Figures 47

List of Tables 48

DIgSILENT PowerFactory 2024, Technical Reference ii

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
1 General Description

1 General Description

This document describes the models of transmission lines available in PowerFactory.

PowerFactoryprovides models which can represent dc and ac lines for all possible phase tech-
nologies (3-phase, 2-phase and 1-phase; with/without a neutral conductor and ground wires)
for both single- and mutually-coupled parallel circuits. Table 1.1 provides an overview of the
supported options and the corresponding element/type combinations.

Table 1.1: Overview of line models available in PowerFactory

System Phase Technology Element Type

DC Unipolar ElmLne TypLne
AC, single-circuit 1-ph ElmLne TypLne
2-ph ElmLne TypLne
3-ph ElmLne TypLne, TypTow,
TypGeo
1-ph with neutral ElmLne TypLne
2-ph with neutral ElmLne TypLne
3-ph with neutral ElmLne TypLne
AC, mutually-coupled circuits Any combination of ElmTow TypTow, TypGeo
phase technologies
DC, mutually-coupled circuits Any combination of ElmTow TypTow, TypGeo
phase technologies
AC/DC, mutually-coupled cir- Any combination of ElmTow TypTow
cuits phase technologies

The line element ElmLne in PowerFactory is the constituent element of transmission lines.
When referring to a type, the line element can be used to define single-circuit lines of any
phase technology according to Table 1.1. In addition, the element parameter Number of Parallel
Lines allows the representation of parallel lines without mutual coupling.

1.1 Mutual Coupling

If the mutual coupling between parallel lines is to be considered, then a line coupling element
ElmTow has to be defined. In this case, the line element ElmLne points to a line coupling ele-
ment ElmTow which in turns refers to the corresponding tower type (TypTow) or tower geometry
type (TypGeo).

Alternatively zero sequence mutual impedance objects (ElmZ0mut) can be used to define zero
sequence coupling impedances of three phase lumped parameter line models. For details, see
section 4.1.

1.1.1 AC/DC Couplings

AC and DC circuits increasingly share rights-of-way, often with only small distances between
the transmission lines, resulting in electromagnetic coupling effects. Voltage and current can
be induced on either AC or DC lines under both steady-state and transient conditions. AC/DC
tower couplings can be modelled by using the line coupling element (ElmTow) in conjunction
with the tower type (TypTow). The tower type must have the System Type set to AC/DC. The
impedances and admittances of the AC circuit/s will be solved using the user-defined Nominal

DIgSILENT PowerFactory 2024, Technical Reference 1

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
1 General Description

Frequency, and of the DC circuit/s using the Nominal Frequency (DC). AC/DC couplings are
handled as follows in the various PowerFactory calculations:

• Load Flow: The transmission line equations are solved separately for the AC circuit/s
and DC circuit/s. The DC circuit/s are always solved using the lumped parameter model,
regardless of the selection of the Line Model in the line coupling element (ElmTow).
• Short-Circuit: AC circuit/s are considered and any DC circuit/s are ignored.
• RMS Simulation: The DC circuit/s are always solved using the lumped parameter model,
regardless of the selection of the Line Model in the line coupling element (ElmTow).
• EMT Simulation: The frequency dependent phase-domain model is selected by default
in the line coupling element (ElmTow). The user can choose between the Universal Line
Model (ULM) or the Frequency Dependent Model (FDM).

1.2 Frequency Dependency

PowerFactoryfurther distinguishes between constant and frequency dependent parameter mod-

els. Models based on tower geometry types (TypTow or TypGeo) use frequency dependent
parameters. This means that the electrical parameters of the line per unit length are calculated
from the mechanical characteristics of the tower and the conductors accounting for skin effect,
the frequency dependent earth-return path of the line, etc. These types should be selected for
use in simulations where a wide range of frequencies is involved or frequencies other than the
nominal system frequency. For further information about the calculation of the per unit length
parameters, please refer to [1].

Models based on line types (TypLne) are by default not frequency dependent. The user enters
the electrical parameters per unit length of the line at system frequency. These parameters
remain unchanged; if the frequency of the simulation changes, i.e. differs from the system
frequency, then the program will adjust the reactance and susceptance of the line according to
the new frequency but the inductances and capacitances will remain unchanged. For certain
calculations (i.e. harmonic load flow, frequency sweep) the user still has the option of assigning
a frequency characteristic to the parameters in the line type. Further details pertaining to input
parameters for the different phase technologies (3-phase, 2-phase, 1-phase, w/o neutral) and
frequency characteristics are discussed in the following sections.

1.3 Lumped and Distributed Parameters

For three-phase lines (either single or multiple parallel circuits), the user can choose between
lumped or distributed parameters. For long transmission lines the distributed parameter model
is preferred as it gives highly accurate results, while the lumped parameter model provides
sufficient results for short lines. The details of the different models are discussed in the following
sections.

Rating of the line/cable

The rated current is the base for per-unit current quantities. The rated current Ir of the three-
phase line/cable is calculated as:
Ir = sline · nlnum kA (1)
or for a cable with a selected line type (TypLne) and the option Air for laying is selected:
Ir = InomAir · nlnum kA (2)
where:

DIgSILENT PowerFactory 2024, Technical Reference 2

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
1 General Description

• sline is the rated current in kA from the line type (TypLne, sline). In case of using a tower
type or a tower geometry the rated current (sline) from the corresponding conductor type
(TypCon) is used. Note the rated current of the conductor type is the total rated current of
the whole bundle.
• nlnum is the number of parallel lines from

• InomAir is the rated current in air in kA from the line type (TypLne)

If line sections are defined in the line/cable, the smallest rated current of all sections is used as
the rated current.

1.4 Line Shunt Compensation

Shunt reactors are used to compensate reactive power generation in long transmission lines,
thereby increasing the efficiency and power quality of the network. Transmission line elements
(ElmLne) in PowerFactory can be defined to have shunt compensation at each end of the line
via the use of Line Compensation elements (ElmLnecomp). Such elements are accessible
either via the Sections/Line Loads/Compensation button on the Basic Data page or via the
Compensation subpage. In addition, the Project Settings (Calculation Options → Lines page)
provide an option to consider the line compensation current when calculating the line loading.

1.4.1 Input Mode

The two input modes and associated input parameters available in the Line Compensation ele-
ment (ElmLnecomp) are described below.

1.4.1.1 Power values

Input fields are available for P 1 and P 0 in MW, and Q1 and Q0 in Mvar. The Compensation type
is automatically adjusted to R-L or C based on the sign of Q1, as illustrated in Figure 1.1.

Figure 1.1: Positive sequence input parameters

DIgSILENT PowerFactory 2024, Technical Reference 3

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
1 General Description

1.4.1.2 Design parameters

Input fields are available for the Reactive power Qrea and Quality factor at fn qf rea, R0/R1 ratio
R0/R1 and X0/X1 ratio, X0/X1.

The conversion from Reactive power, Qrea , and Quality factor to P 1 and Q1 is performed ac-
cording to:

Qrea · (1/qf rea + j)

P 1 + j · Q1 =
(1/qf rea )2 + 1
Qrea · 1/qf rea
P1 = (3)
(1/qf rea )2 + 1
Qrea
Q1 =
(1/qf rea )2 + 1

and vice versa:

Qrea = Q1 · (1/qf rea )2 + 1)
Q1 (4)
qf rea =
P1
If qf rea = 0, then P 1 = 0 and Q1 = Qrea .

The frequency dependence of the impedance can be defined on the Power Quality/Harmonics
page of the element using a relative characteristic.

DIgSILENT PowerFactory 2024, Technical Reference 4

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

2 Equivalent Circuits

2.1 Lumped Parameters

Figure 2.1 shows the equivalent PI-circuit used by PowerFactory to represent AC transmis-
sion lines with lumped parameters. The subscripts s and r denote the sending and receiving
ends of the line, respectively. The general formulation discussed in this section is valid for any
phase technology by appropriate dimensioning of the impedance and admittance matrices, even
though the formulation presented here is based on a 3-phase line with no neutral conductor.

Figure 2.1: Equivalent PI-circuit of the line for lumped parameters

The equations of the voltages and currents at the sending and receiving ends of the line are
formulated in terms of impedance and admittance matrices. The dimension of the matrices
depends on the phase technology. The longitudinal voltage drop along the line is given by the
impedance matrix in the following form:

         
U s,A U r,A ∆U A Zs Zm Zm IA
 U s,B  −  U r,B  =  ∆U B  =  Z m Zs Zm  ·  IB  (5)
U s,C U r,C ∆U C Zm Zm Zs IC

According to the sign convention assumed in Figure 2.1, the current at the sending end of the
line is calculated in terms of the admittance matrix as follows:

           
I s,A ∆I s,A IA Ys Ym Ym U s,A IA
 I s,B  =  ∆I s,B  +  I B  = 1  Y m Ys Y m  ·  U s,B  +  I B  (6)
2
I s,C ∆I s,C IC Ym Ym Ys U s,C IC

Similarly, the current at the receiving end of the line is given by:

DIgSILENT PowerFactory 2024, Technical Reference 5

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

           
I r,A ∆I r,A IA Ys Ym Ym U r,A IA
1
 I r,B  =  ∆I r,B  −  I B  =  Y m Ys Y m  ·  U r,B  −  I B  (7)
2
I r,C ∆I r,C IC Ym Ym Ys U r,C IC

Equations (5), (6) and (7) completely define the PI-model of the line for lumped parameters.
The impedance and admittance matrices:

   
Zs Zm Zm Ys Ym Ym
[Z ABC ] =  Z m Zs Zm  [Y ABC ] =  Y m Ys Ym  (8)
Zm Zm Zs Ym Ym Ys

are the natural impedance and admittance matrices of the line after reduction of earth wires (if
any).

Note that Y s represents the sum of all admittances connected to the corresponding phase, while
Y m is the negative value of the admittance between two phases. Similarly, Y p is the sum of all
admittances connected to the neutral conductor and Y pn is the negative value of the admittance
between the neutral and the phase conductors.

The PI-circuit described here is the general formulation of the line model with lumped parame-
ters in PowerFactory. The following sections discuss the application of the model to the different
phase technologies (3-, 2-, 1-phase, w/o neutral conductors) and the required user-defined
parameters in each case.

2.1.1 Single-Phase Line

The equivalent circuit in Figure 2.1 can be reduced to the PI-circuit in Figure 2.2 for the single-
phase line.

Figure 2.2: Lumped parameter model for a single-phase AC line

The impedance and admittance of the equivalent circuit are calculated from the input parame-
ters defined in the line type (TypLne) according to the following equations:

0
 0 0

Z = Z1 · l = R1 + jωL1 · l
0
 0 0

Y = Y1 · l = G1 + jωC1 · l (9)
0 0
G1 = B1 · tgδ1

DIgSILENT PowerFactory 2024, Technical Reference 6

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

0 0 0 0
where l is the length of the line in km, and R1 ,L1 , G1 and C1 are the line parameters per unit
0
length. Note that the conductance G1 can be defined in terms of the insulation factor, tgδ. The
reader is referred to Table A.1 for the complete list of input parameters.

The currents and voltages on both sides of the line in Figure 2.2 are related by the following
equation:

     
Us A B Ur
= · (10)
Is C D −Ir

where the ABCD parameters of the equivalent PI-circuit are:

0 0
A = 1 + 21 Z · Y · l2
0
B =Z ·l
 0 0  (11)
0
Z ·Y ·l2
C =Y ·l· 1+ 4

D=A

2.1.2 Single-Phase Line with Neutral Conductor

From Figure 2.1, the equivalent circuit shown in Figure 2.3 can be now deduced for the single-
phase line with neutral conductor model.

Figure 2.3: Equivalent circuit of the single-phase line with neutral conductor model

The self- and mutual impedances and admittances of the equivalent circuit:

   
Zs Z pn Ys Y pn
[Z an ] = [Y an ] = (12)
Z pn Zn Y pn Ys

are calculated from the input parameters defined in the line type (TypLne).

Thus the self and mutual impedances and admittances in (12) are related to the input parame-
ters Z 1 , Z n , Z pn , Y 1 , Y n and Y pn as follows:

Zs = Z 1 Ys = Y 1 (13)

DIgSILENT PowerFactory 2024, Technical Reference 7

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

The self- and mutual impedance Z n , Z pn and the admittance for the neutral conductor Y n , Y pn
are as listed in Table A.1 (Input parameter of the line type (TypLne)) of annex A.

2.1.3 Two-Phase Line

From Figure 2.1, the equivalent circuit shown in Figure 2.4 can be now deduced for the 2-phase
line model.

Figure 2.4: Equivalent circuit of the two-phase line model

The self- and mutual impedances and admittances of the equivalent circuit:

   
Zs Zm Ys Ym
[Z ab ] = [Y ab ] = (14)
Zm Zs Ym Ys

are calculated from the input parameters defined in the line type (TypLne). The input parameters
- positive and zero sequence components - are converted to the impedances and admittances
in (14) via the following transformation:

   
  1 1  −1 1 1 1
T S2ph = T S2ph =
1 −1 2 1 −1

−1
[Z 01 ] = [TS2ph ] × [Z ab ] × [TS2ph ]

Thus the self and mutual impedances and admittances in (14) are related to the input parame-
ters Z 1 , Z 0 , Y 1 and Y 0 as follows:

1 1
Zs = · (Z 0 + Z 1 ) Ys = · (Y 0 + Y 1 ) (15)
2 2

1 1
Zm = · (Z 0 − Z 1 ) Ym = · (Y 0 − Y 1 ) (16)
2 2

or can be resolved in terms of the sequence magnitudes:

DIgSILENT PowerFactory 2024, Technical Reference 8

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

   
Z0 0 Zs + Zm 0
[Z 01 ] = =
0 Z1 0 Zs − Zm

   
Y0 0 Ys+Ym 0
[Y 01 ] = =
0 Y1 0 Ys−Ym

2.1.4 Two-Phase Line with Neutral Conductor

From Figure 2.1, the equivalent circuit shown in Figure 2.5 can be now deduced for the 2-phase
line with neutral conductor model.

Figure 2.5: Equivalent circuit of the two-phase line with neutral conductor model

The self- and mutual impedances and admittances of the equivalent circuit are:

   
Zs Zm Z pn Ys Ym Y pn
[Z abn ] =  Z m Zs Z pn  [Y abn ] =  Y m Ys Y pn  (17)
Z pn Z pn Zn Y pn Y pn Yn

The input parameters of the model are the positive and zero sequence impedances Z 1 , Z 0 . The
positive and zero sequence admittance, Y 1 and Y 0 , the self- and mutual impedance Z n , Z pn
and the admittance for the neutral conductor Y n , Y pn are as listed in Table A.1 (Input parameter
of the line type (TypLne)) of annex A.

The values Z n , Z pn , Y p and Y pn of the neutral conductor can be directly used in (17).

Thus the self and mutual impedances and admittances in (17) are related to the input param-
eters Z 1 , Z 0 , Y 1 and Y 0 and are determined like for the two phase line model (see equations
(15) and (16)).

2.1.5 Three-Phase Lines

The equivalent circuit of the three-phase line is shown in Figure 2.6. The self- and mutual
impedances and admittances are given by:

DIgSILENT PowerFactory 2024, Technical Reference 9

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

   
Zs Zm Zm Ys Ym Ym
[Z abc ] =  Z m Zs Zm  [Y abc ] =  Y m Ys Ym  (18)
Zm Zm Zs Ym Ym Ys

Figure 2.6: Equivalent circuit of the three-phase line

The input parameters in the line type (TypLne) are defined in terms of positive and zero se-
quence impedances and admittances Z 1 , Y 1 , Z 0 and Y 0 . The negative sequence is assumed
to be equal to the positive sequence.

The conversion from the sequence components into the natural components in (18) is done via
the complex transformation matrix [Ts ] as follows:

   
1 1 1 1 1 1
−1 1
[T S ] =  1 a2 a  → [T S ] =  1 a a2  (19)
3
1 a a2 1 a2 a

−1
[Z 012 ] = [TS ] × [Z abc ] × [TS ]

Thus the self and mutual impedances and admittances in (18) are related to the input parame-
ters Z 1 , Z 0 , Y 1 and Y 0 as follows:

1 1
Zs = · (Z 0 + 2 · Z 1 ) Ys = · (Y 0 + 2 · Y 1 ) (20)
3 3

1 1
Zm = · (Z 0 − Z 1 ) Ym = · (Y 0 − Y 1 ) (21)
3 3

or can be resolved in terms of the sequence magnitudes:

   
Z0 0 0 Z s + 2Z m 0 0
[Z 012 ] =  0 Z1 0 = 0 Zs − Zm 0 
0 0 Z2 0 0 Zs − Zm

DIgSILENT PowerFactory 2024, Technical Reference 10

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

   
Y0 0 0 Y s + 2Y m 0 0
[Y 012 ] =  0 Y1 0 = 0 Ys−Ym 0 
0 0 Y2 0 0 Ys−Ym

2.1.6 Three-Phase Line with Neutral Conductor

Figure 2.7 shows the equivalent circuit of the 3-phase line with neutral conductor. The voltages
and the currents at both ends of the line are related by the impedance and admittance matrices:

   
Zs Zm Zm Z pn Ys Ym Ym Y pn
 Zm Zs Zm Z pn   Ym Ys Ym Y pn 
[Z abcn ] = 
 Zm
 [Y abcn ] =   (22)
Zm Zs Z pn   Ym Ym Ys Y pn 
Z pn Z pn Z pn Zn Y pn Y pn Y pn Yn

Figure 2.7: Equivalent circuit for the 3-phase line with neutral conductor

The input parameters of the model are the positive and zero sequence impedances Z 1 , Z 0 . The
positive and zero sequence admittance, Y 1 and Y 0 , the self- and mutual impedance Z n , Z pn
and the admittance for the neutral conductor Y n , Y pn are as listed in Table A.1 (Input parameter
of the line type (TypLne)) of annex A.

The values Z n , Z pn , Y p and Y pn of the neutral conductor can be directly used in (22). The self-
and mutual impedance are determined like for the three phase line model (see equations (20)
and (21)).

2.1.7 4-Wire Model Data Conversion

The self- and mutual impedances and admittances are not always available in the format re-
quired by the line type (TypLne). The following subsections guide the reader on how to convert
commonly available measurement data into the required input parameters.

DIgSILENT PowerFactory 2024, Technical Reference 11

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

2.1.7.1 Measurement between phase A and phase B wire

Figure 2.8: Phase to phase measurement loop

As per the impedance matrix in (22):

     
Ua Zs Zm Zm Z pn Ia
 U b   Zm Zs Zm Z pn   Ib 
 U c  =  Zm
      (23)
Zm Zs Z pn   I c 
Un Z pn Z pn Z pn Zn In

thus:
U a = Zs · Ia + Zm · Ib
U b = Zm · Ia + Zs · Ib

According to the measurement in Figure 2.8:

I a = (−I b )
Ua − Ub = 2Ia · (Z s − Z m )

the positive sequence impedance can be derived as follows:

1 Ua − Ub
Z 1 = R1 + jX1 = (24)
2 Ia

DIgSILENT PowerFactory 2024, Technical Reference 12

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

2.1.7.2 Measurement between neutral and PE (earth) wire

Figure 2.9: Neutral to ground measurement loop

Again from (23) now with Ia=Ib=Ic=0:

Un = Z n · I n

Un
Zn = (25)
In

Note that Z n represents the impedance of the neutral conductor, Z N eutral , plus the impedance
of the earth-return path, Z earth .

2.1.7.3 Measurement between phase and PE (earth) wire

Figure 2.10: Phase – PE (earth wire) measurement

From equation (23) with In = 0

1
U= (Z · I + Z m · I + Z m · I)
3 s

DIgSILENT PowerFactory 2024, Technical Reference 13

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

 
1
U= I · Z0
3

the zero-sequence impedance is:

3·U
Z0 = (26)
I

2.1.7.4 Measurement between phase and neutral wire

Figure 2.11: Phase – neutral measurement

From equation (23):

1
U= (Z · I + Z m · I + Z m · I) + Z pn · I n
3 s
1

U n = · I · Z pn + Z pn + Z pn + Z n · I n
3

With I n = −I:
1
U= (Z · I + Z m · I + Z m · I) − Z pn · I
3 s
1

U n = · I · Z pn + Z pn + Z pn − Z n · I
3

And according to equations (20) and (21):

1
U= Z · I − Z pn · I
3 0

U n = I · Z pn − Z n

Subtracting these equations:

3 (U − U n )
Z 0,P H−N = = Z 0 − 6 · Z pn + 3 · Z n
I

DIgSILENT PowerFactory 2024, Technical Reference 14

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

Z 0,P H−N is commonly referred to as the zero-sequence impedance between with return over
the neutral conductor. With Z n given by (25) the mutual impedance between phase and neutral
conductors is:

Z 0 + 3 · Z n − Z 0,P H−N
Z pn = (27)
6

2.1.7.5 Data conversion without N-PE measurement

If the measurement between the neutral and the PE (earth) wire does not exist the following
simplification can be made:

Phase-neutral loop:
Z 0,P H−N = Z 1 + 3 · Z N eutral (28)

Phase-ground loop:
Z 0 = Z 1 + 3 · Z Earth (29)

Thus the neutral-earth loop impedance:

Z N −E = Z N eutral + Z Earth

Z 0,P H−N + Z 0 − 2 · Z 1
Z N −E = Z n = (30)
3

And the phase-neutral mutual impedance:

Z0 − Z1
Z pn = (31)
3

2.2 Distributed Parameter Model

In addition to the lumped parameter models described in previous sections, PowerFactory also
supports distributed parameters models for 3-phase line circuits. This kind of model accounts
for the distributed nature of the line parameters and should therefore be the preferred option
for the modelling of long lines. For short lines, the lumped parameter models discussed in the
previous sections should provide sufficiently accurate solutions.

A line is considered to be long when its physical length is of the same order of magnitude as
the length of wave of the voltage/current at the frequency under consideration (e.g. system
frequency for the Load Flow calculation). Note that for increasing frequencies, and hence for
example for harmonic Load Flow calculations, the higher the frequency the lower the length of
wave, so that even a physically short line may need to be treated as a long line, and therefore
be represented using distributed parameters.

To select a distributed parameter model, the corresponding model option on the Basic Data
page of the line element (ElmLne) or line coupling (ElmTow) should be selected.

DIgSILENT PowerFactory 2024, Technical Reference 15

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

2.2.1 General Formulation

Equations (32) and (33) describe the incremental transmission line model in the frequency do-
main of an elemental length ∆x depicted in Figure 2.12.

∂
V = I(x) · Z 0 (32)
∂x

∂
I = V (x) · Y 0 (33)
∂x

Figure 2.12: Incremental model for a line of elemental length

After taking the second derivatives of (32) and (33) with respect to x and rearranging the equa-
tions to separate the voltage from the current magnitudes, the system of differential equations
can be rewritten as:

∂2
∂x2 V = Z 0 · Y 0 · V (x)
(34)
∂2 0 0
∂x2 I = Z · Y · I(x)

The general solution is of the form:

U (x) = K1 · eγ·x + K2 · e−γ·x

(35)
ZC · I(x) = −K1 · eγ·x + K2 · e−γ·x

with r
Z0
ZC = (36)
Y0

√
γ= Z 0 · Y 0 = α + jβ (37)

DIgSILENT PowerFactory 2024, Technical Reference 16

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

Both the surge (or characteristic) impedance ZC and the propagation factor γ are frequency-
dependent and uniquely characterise the behaviour of the transmission line. Further details
regarding the derivation of these equations can be found in [1], [3].

The integration constants K1 and K2 in (35) are determined from the border conditions at either
the receiving or the sending end of the line. According the sign convention in Figure 2.13, the
particular solution of (35) results:

         
Ur cosh γ · l −ZC · sinh γ · l Us A B Us
= 1 · = · (38)
Ir ZC · sinh γ · l − cosh γ · l Is C D Is

and therefore the impedance and admittance of the equivalent circuit are:
sinh γ·l
Z = ZC · sinh γ · l = Z 0 · l · γ·l
(39)
cosh γ·l−1 tanh( γ·l
2 )
Y = ZC ·sinh γ·l = 1
2 ·Y0·l· γ·l
2

Figure 2.13: Equivalent pi-circuit for the line with distributed parameters in the frequency domain

It should be noted that Z and Y in Figure 2.13 are frequency-dependent parameters as both
the surge impedance, ZC , and the propagation factor, γ, are functions of the frequency.

Series expansion

The lumped parameter model described by (11) in Section 2.1.1 is a simplified model of the
distributed parameters model. Taking a series expansion of the hyperbolic functions in (39)
gives:
1 1 1
cosh ϑ = 1 + · ϑ2 + · ϑ4 + · ϑ6 + . . .
2 24 720

sinh ϑ 1 1 1
= 1 + · ϑ2 + · ϑ4 + · ϑ6 + . . .
ϑ 6 120 5040
√
Using ϑ = γ · l = Z 0 · Y 0 · l A and B in (38) can be expanded as follows:

1 1 2
A = cosh γ · l = 1 + · Z 0 · Y 0 · l2 + · (Z 0 · Y 0 ) · l4 + . . .
2 24

   
0 sinh γ · l 0 1 0 0 2 1 0 0 2 4
B =Z ·l· =Z ·l· 1+ ·Z ·Y ·l + · (Z · Y ) · l + . . .
γ·l 6 120

DIgSILENT PowerFactory 2024, Technical Reference 17

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
2 Equivalent Circuits

Considering up to the second order terms, equations (39) of the distributed parameter model
go into equations (9) of the lumped parameter model:

Z = B = Z 0 · l = R0 · l + jω · L0 · l

A−1 1+ 1
· Z 0 · Y 0 · l2 1 1
Y = = 2
0
= · Y 0 · l = · (G0 · l + jω · C 0 · l)
B Z ·l 2 2

The accuracy of the lumped model therefore depends on the weight of truncated terms in the
series expansion, which in turns depends on the factor f · l (frequency multiplied by length).
For overhead lines with a length less than 250 km and at system frequency, this approximation
is sufficient and the error is negligible. For longer lines or higher frequencies, a distributed
parameter model will provide a more accurate solution.

Longer lines can be alternatively modelled by cascading line sections. In general, the longer
the line or the higher the frequency, the more line sections are required in order to obtain the
same accuracy. Increasing the number of line sections to infinity will turn the lumped parameter
model into the distributed parameters model discussed previously.

DIgSILENT PowerFactory 2024, Technical Reference 18

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
3 Load Flow Analysis

3 Load Flow Analysis

3.1 Temperature Dependency of Line Resistance

For a line type a temperature dependency can be defined either via the conductor material:

• Aluminium (α = 0.00403 1/K)

• Copper (α = 0.00393 1/K)
• Aldrey (α = 0.00360 1/K)
• Aluminium-Steel (α = 0.00403 1/K)
• Aldrey-Steel (α = 0.00360 1/K)

with a temperature coefficient α or with a resistance at max. operating temperature (tmax in

degC and rlinetmax in Ohm/km)

Depending on the temperature setting e.g. in the load flow command, frequency sweep the
positive sequence resistance is scaled as follow when using a line type (TypLne):
0
R1 = rline · (1 + α · (temp − 20) (40)

Note: The zero sequence resistance is not adapted.

At which temperature the resistance is modified depends on the setting e.g. of the load flow
command or/and e.g. the entered operating temperature at the line.

• at 20 degC
• at max. operating temperature (tmax in line type, conductor type, cable type)
• at operating temperature (T op entered in the line model)
• at temperature entered in load flow or frequency sweep command

Note: When using a tower model, and the temperature is configured as ”at operating tempera-
ture”, the earth conductor is still applied by the entered resistance at 20 degC. For lines using a
cable type (TypCabsys), or a cable system (TypCabsys), the conductor, sheath and armour the
corresponding temperature is applied.

3.2 AC Model Calculation Parameters

Loading

The loading of the line/cable is calculated as follows:

 
|Ibusi | |Ibusj |
loading = max , · 100 % (41)
Inom(busi) Inom(busj)

where:

• Inom(busi) is nominal current of the line/cable for terminal i in kA

DIgSILENT PowerFactory 2024, Technical Reference 19

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
3 Load Flow Analysis

• Inom(busj) is nominal current of the line/cable for terminal j in kA

• Ibusi is magnitude of the current at terminal i

• Ibusj is magnitude of the current at terminal j

If no thermal ratings object is defined, the nominal current is:

0 0
Inom(busi) = Inom(busj) = Ir · f line (see equation 1 or 2).

where f line is the derating factor of the line (if a thermal ratings object is selected, f line is not
used in the calculation).

If a thermal ratings object is used, the nominal currents are calculated using the parameter
ContRating and Un(busi) , Un(busj) (nominal voltage in kV of the connected terminal i and j) as:

• if the continuous rating is entered in MVA:

– 3-phase AC:
√ 
0
Inom(busi) = ContRating/ 3 · Un(busi) · nlnum kA
√ 
0
Inom(busj) = ContRating/ 3 · Un(busj) · nlnum kA

– 2-phase AC:
0
Inom(busi) = ContRating/Un(busi) · nlnum kA
0
Inom(busj) = ContRating/Un(busj) · nlnum kA

– 1-phase AC:
0
√
Inom(busi) = ContRating · 3/Un(busi) · nlnum kA
0
√
Inom(busj) = ContRating · 3/Un(busj) · nlnum kA

– 1-phase connected on a AC/BI system:

0
Inom(busi) = ContRating · 2/Un(busi) · nlnum kA
0
Inom(busj) = ContRating · 2/Un(busj) · nlnum kA

• if the continuous rating is entered in kA:

0 0
Inom(busi) = Inom(busi) = ContRating · nlnum kA

• if the continuous rating is entered in %:

0 0
Inom(busi) = Inom(busi) = ContRating/100 · Ir kA

0 0
The nominal currents are generally the same (Inom(busi) = Inom(busj) ), but not if the line/cable
is connected between different voltage levels (Un(busi) 6= Un(busi) ).

Further details regarding thermal ratings object (IntThrating), see Technical Reference of the
Thermal Ratings Definition

Note: In case if the nominal current is greater than the max. permissible current, the max.
permissible current (Imaxlim in kA, see Load Flow page of the line) is used for the nominal
current.

DIgSILENT PowerFactory 2024, Technical Reference 20

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
3 Load Flow Analysis

0
• Inom(busi) = min(Inom(busi) , Imaxlim · nlnum)
0
• Inom(busj) = min(Inom(busj) , Imaxlim · nlnum)

For an unbalanced load flow calculation the highest current of all phases is used.

Losses

The losses are calculated as:

Losses = (Pbusi + Pbusj ) · 1000 kW (42)
where Pbusi and Pbusj are calculation parameters available for the branch element.

The total losses are obtained as:

P loss = Pbusi + Pbusj − Plineloads MW
(43)
Qloss = Qbusi + Qbusj − Qlineloads Mvar
where Plineloads and Qlineloads are the active and reactive power of the line loads.

The load losses are obtained using the no-load loss parameters:
P lossld = P loss − P lossnld MW
(44)
Qlossld = Qloss − Qlossnld Mvar

The no-load losses are calculated using the Active no-load losses Gload and the Capacitative
loading Cload as follows:
S lossnld = ubusi · (y line · ubusi )∗ + u0busi · (y line · u0busj )∗ MVA
Gload = <(S lossnld ) · 1000 kW
Cload = −=(S lossnld ) Mvar (45)
P lossnld = Gload/1000 MW
Qlossnld = −Cload Mvar
where ubusi and ubusj are the terminal voltages where the line is connected in p.u., u0busj =
ubusj · Un(busj) /Un(busi) (nominal voltages of the connected busbars) and y line is the complex
admittance of the line.

Voltage Drop

The voltage drop and voltage angle drops are calculated as:
du = |ubusi | − |ubusj | p.u.
dupc = du · 100 % (46)
dphiu = φu busi − φu busj deg
and the positive-sequence voltage drop and positive-sequence voltage angle drop are calcu-
lated as:
d1u = |u1busi | − |u1busj | p.u.
du1pc = du1 · 100 % (47)
dphiu1 = φu1,busi − φu1,busj deg
where ubusi and ubusj are the amplitudes of the corresponding terminal voltage in p.u. based on
the rated voltage of the terminal, φu busi and φu busj are the terminal voltage angles in deg. For
an unbalanced load flow du, dupc, and dphiu are available per phase (e.g. c : dupc : B).

DIgSILENT PowerFactory 2024, Technical Reference 21

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
3 Load Flow Analysis

3.3 AC Model Calculation Parameters for Linear DC Load Flow

Loading

The loading of the line/cable is calculated as follows:

 
|Pbusi | |Pbusj |
loading = max , · 100 %
Pnom(busi) Pnom(busj)
where:

• Pbusi : Active power at terminal i

• Pbusj : Active power at terminal j
• Pnom(busi) : Nominal power at terminal i
• Pnom(busj) : Nominal power at terminal j

The nominal power is determined as follow when no thermal ratings object is defined:

0
√
• Pnom(busi) = 3 · Un(busi) · Ir · f line
0
√
• Pnom(busj) = 3 · Un(busj) · Ir · f line

with the rated current of the line/cable Ir (see equation 1 or 2).

where f line is the derating factor of the line (if a thermal ratings object is selected, f line is not
used in the calculation).

Un(busi) , Un(busj) is the nominal voltage in kV of the connected terminals.

If a thermal ratings object is used, the nominal power is calculated using the parameter ContRating
and Un(busi) and Un(busj) (nominal voltage in kV of the connected terminal i and j) as:

• if the continuous rating is entered in MVA:

0 0
Pnom(busi) = Pnom(busj) = ContRating · nlnum MW

• if the continuous rating is entered in kA:

0
√
Pnom(busi) = 3 · Un(busi) · ContRating · nlnum MW
0
√
Pnom(busj) = 3 · Un(busj) · ContRating · nlnum MW

• if the continuous rating is entered in %:

0
√
Pnom(busi) = 3 · Un(busi) · ContRating/100 · Ir MW
0
√
Pnom(busj) = 3 · Un(busj) · ContRating/100 · Ir MW

0 0
The nominal powers are generally the same (Pnom(busi) = Pnom(busj) ), but not if the line/cable
is connected between different voltage levels (Un(busi) 6= Un(busi) ).

Further details regarding thermal ratings object (IntThrating), see Technical Reference of the
Thermal Ratings Definition

Note: In case if the nominal power is greater than the max. permissible power (based on the
parameter Imaxlim in kA, see Load Flow page of the line), the max. permissible power is used
for the nominal power:

DIgSILENT PowerFactory 2024, Technical Reference 22

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
3 Load Flow Analysis

0
√
• Pnom(busi) = min(Pnom(busi) , 3 · Un(busi) · Imaxlim · nlnum)
0
√
• Pnom(busj) = min(Pnom(busj) , 3 · Un(busj) · Imaxlim · nlnum)

Losses

Losses are not calculated in the linear DC Load Flow.

3.4 DC Model Calculation Parameters

Loading

The loading of the line/cable is calculated as follows:

 
|Ibusi | |Ibusj |
loading = max , · 100 %
Inom(busi) Inom(busj)

• Inom(busi) is nominal current of the line/cable for terminal i in kA

• Inom(busj) is nominal current of the line/cable for terminal j in kA
• Ibusi DC current at terminal i
• Ibusj DC current at terminal j

If no thermal ratings object is defined, the nominal current is:

0 0
Inom(busi) = Inom(busj) = Ir · f line (see equation 1 or 2).

where f line is the derating factor of the line (if a thermal ratings object is selected, f line is not
used in the calculation).

If a thermal ratings object is used, the nominal currents are calculated using the parameter
ContRating and Un(busi) ,Un(busj) (nominal voltage in kV of the connected terminal i and j) as:

• if the continuous rating is entered in MVA:

0
Inom(busi) = ContRating/Un(busi) · nlnum kA
0
Inom(busj) = ContRating/Un(busj) · nlnum kA

• if the continuous rating is entered in kA:

0 0
Inom(busi) = Inom(busj) = ContRating · nlnum kA

• if the continuous rating is entered in %:

0 0
Inom(busi) = Inom(busj) = ContRating/100 · Ir kA

0 0
The nominal currents are generally the same (Inom(busi) = Inom(busj) ), but not if the line/cable
is connected between different voltage levels (Un(busi) 6= Un(busi) ).

Further details regarding thermal ratings object (IntThrating), see Technical Reference of the
Thermal Ratings Definition

Note: In case if the nominal current is greater than the max. permissible current, the max.
permissible current (Imaxlim in kA, see Load Flow page of the line) is used for the nominal
current.

DIgSILENT PowerFactory 2024, Technical Reference 23

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
3 Load Flow Analysis

0
• Inom(busi) = min(Inom(busi) , Imaxlim · nlnum)
0
• Inom(busj) = min(Inom(busj) , Imaxlim · nlnum)

Losses

The losses are calculated as:

Losses = (Pbusi + Pbusj ) · 1000 kW (48)

where Pbusi and Pbusj are calculation parameters available for the branch element.

The total losses are obtained as:

P loss = Pbusi + Pbusj MW
(49)
Qloss = 0 Mvar

The load losses are obtained using the no-load loss parameters:

P lossld = P loss − P lossnld MW

(50)
Qlossld = 0 Mvar

The no-load losses are calculated using the Active no-load losses Gload as follows:
2 2
Gload = (1/2 · glne · Ubusi + 1/2 · glne · Ubusj ) · 1000 kW
Cload = 0 Mvar
(51)
P lossnld = Gload/1000 MW
Qlossnld = 0 Mvar

where glne is the admittance (calculation parameter) of the line and Ubusi and Ubusj are the
terminal DC voltages where the line is connected.

3.5 Consideration of Line Shunt Compensation

For load flow calculations, line shunt compensation is modelled as a shunt admittance at each
end, i,j, of the line: y i and y j .
comp comp

These shunt admittances are used to calculate the shunt compensation currents, icomp , as
follows:
icompi = ii − y i · ui
comp
(52)
icompj = ij − y j · uj
comp

For unbalanced load flow calculations, the voltages are first transformed to the sequence system
and then multiplied by the positive and zero sequence admittances accordingly. Transformation
back to the phase domain is then used to obtain the phase shunt compensation currents.

DIgSILENT PowerFactory 2024, Technical Reference 24

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
4 Short-Circuit

4 Short-Circuit

For short-circuit calculations, transmission lines can be represented using either the lumped
parameter model or the distributed parameter model. These are described in Sections 2.1 and
2.2, respectively.

Note: The distributed parameter model is not used by the VDE and ANSI methods. The Com-
plete method will only use the distributed parameter model if the Short-Circuit command options
are configured as follows:

• Load flow initialisation is selected; or

• Ignore positive sequence data → Capacitance of lines is deselected

In addition, the distributed parameter line will not be used during the R/X calculation.

4.1 Zero sequence mutual impedance

With the zero sequence mutual impedance objects two or more lines can be coupled to defined
the zero sequence coupling impedances between lines and is only relevant for unbalanced
calculation method.

Note: The zero sequence mutual impedance is only supported for three phase lines or line
sections (without neutral wire).

The zero sequence mutual impedance object an be defined for lines (ElmLne) or line sections
(ElmLnesec). In case if a line is select also the from, to length can be defined. An example is
shown in the following Figure 4.1.

Figure 4.1: Example for zero sequence mutual impedance coupling

For the shown example are three zero sequence mutual impedance objects required z0m1 ,z0m2
and z0m3 .

• z0m1 is configured as follow:

– Line 1: Line A; Start position = 0 %; End position = 40 %

DIgSILENT PowerFactory 2024, Technical Reference 25

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
4 Short-Circuit

– Line 2: Line A; Start position = 0 %; End position = 30 %

– Orientation = same direction
• z0m2 is configured as follow:
– Line 1: Line A; Start position = 60 %; End position = 100 %
– Line 2: Line C; Start position = 0 %; End position = 40 %
– Orientation = opposite direction
• z0m3 is configured as follow:
– Line 1: Line B; Start position = 30 %; End position = 100 %
– Line 2: Line C; Start position = 30 %; End position = 100 %
– Orientation = same direction

The zero sequence impedance is entered in Ω, the from and to position in percent of the line
length. Further the orientation can be defined e.g. in case if two lines are coupled in opposite
direction (see Figure 4.2).

Figure 4.2: Orientation of zero sequence mutual impedance

All zero sequence mutual impedance object will be collected and a zero sequence admittance
matrix will be created

 
Z011 Z0m12 Z0m13 ... Z0m1n

 Z0m21 Z022 Z0m23 ... Z0m2n 

|Z0| = 
 Z0m31 Z0m32 Z033 ... Z0m3n 
 (53)
 ... ... ... ... ... 
Z0mn1 Z0mn2 Z0n3 ... Z0nn

where

• Z011 ,Z022 , Z0nn : zero sequence impedance of the corresponding line 1,2, ... n

• Z0m12 , Z0m13 , Z0m32 ...: zero sequence mutual impedance of the corresponding cou-
pling objects

The zero sequence impedance matrix will be inverted to get a zero sequence admittance matrix:

|Y 0| = |Z0|−1 (54)

DIgSILENT PowerFactory 2024, Technical Reference 26

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
4 Short-Circuit

The zero sequence current for a line is than calculated as follow:

I0 = |Y 0| · |dU 0| (55)

where dU 0 is the zero sequence voltage difference of the coupled line. e.g. for line 1

dU 01 = U 0i,1 − U 0j,1 in case of same direction orientation

dU 01 = U 0j,1 − U 0i,1 in case of opposite direction orientation

In case of a short-circuit at the line, the corresponding zero sequence mutual impedance objects
are depending on the short-circuit location internally split. The zero sequence current of side i
and j are than calculated accordingly.

Figure 4.3: Short-circuit at line with zero sequence mutual impedance

The zero sequence mutual impedance is also split according to the short-circuit location into
two mutual coupling objects.

Note: Only the short-circuit line is split into two PI-Models.

4.2 Consideration of Line Shunt Compensation

Line shunt compensation is ignored by the balanced IEC and ANSI methods, and is also ig-
nored for the R/X ratio calculation (also complete method). When using the Complete method,
the positive line shunt admittance is considered as in the load flow or RMS simulation (see
Section 3.5).

For unbalanced short-circuit calculations, only the zero sequence admittance is considered by
the IEC and ANSI methods. When the R/X ratio is calculated (for the Complete method as well)
the imaginary part of the zero sequence admittance is ignored.

DIgSILENT PowerFactory 2024, Technical Reference 27

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
5 Harmonics Analysis

5 Harmonics Analysis

For harmonics analysis, transmission lines can be represented using either the lumped param-
eter model or the distributed parameter model. These are described in Sections 2.1 and 2.2,
respectively.

5.1 Zero sequence mutual impedance

In case of a zero sequence mutual impedance is defined (see chapter 4.1).

The zero sequence mutual reactance is scaled by the harmonic order.

5.2 Consideration of Line Shunt Compensation

The line compensation is similar as in the EMT simulation considered depending on if the com-
pensation is a reactor or capacitor.

5.2.0.1 Shunt Reactor

For line compensation that is modelled as a shunt reactor, the frequency dependency is applied
as follows:
1
Z 1comp =
Y 1comp
(56)
1
Z 0comp =
Y 0comp

=(Z 1comp ) = =(Z 1comp ) · fh

(57)
=(Z 0comp ) = =(Z 0comp ) · fh

where fh is the harmonic order.

5.2.0.2 Shunt Capacitor

For line compensation that is modelled as a shunt capacitor, the frequency dependency is ap-
plied as follows:

=(Y 1comp ) = =(Y 1comp ) · fh

(58)
=(Y 0comp ) = =(Y 0comp ) · fh

where fh is the harmonic order.

DIgSILENT PowerFactory 2024, Technical Reference 28

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
6 RMS Simulation

6 RMS Simulation

6.1 Zero sequence mutual impedance

In case of a zero sequence mutual impedance is defined (see chapter 4.1).

6.2 Consideration of Line Shunt Compensation

Consideration of line shunt compensation in RMS simulation is as described for the load flow
calculation. Please refer to Section 3.5.

DIgSILENT PowerFactory 2024, Technical Reference 29

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

7 EMT Simulation

The models described in the previous sections are defined in the frequency domain and used
in PowerFactory for all steady-state calculations such as load flow, short-circuit, harmonic load
flow, frequency sweep, and the electromechanical (RMS) simulation.

This section introduces the models used in the electromagnetic transients (EMT) simulation.
These models are obtained via the conversion of frequency-domain models into time-domain
models.

7.1 Lumped Parameter Model

The lumped parameter model discussed in Section 2.1 can be directly used in EMT simula-
tions by replacing the jω operator by d/dt and thus the impedances and admittances by the
corresponding inductances and susceptances.

For the pi-equivalent circuit in Figure 2.1, equations (5), (6) and (7) then become:

     
us,A ur,A ∆uA
 us,B  −  ur,B  =  ∆uB  =
us,C ur,C ∆uC
         
Rs Rm Rm iA Ls Lm Lm iA u0
d
 Rm Rs Rm  ·  iB  +  Lm Ls Lm  ·  iB  +  u0  (59)
dt
Rm Rm Rs iC Lm Lm Ls iC u0

         
∆is,A Gs Gm Gm us,A Cs Cm Cm us,A
1 d
 ∆is,B  =  Gm Gs Gm  ·  us,B  +  Cm Cs Cm  ·  us,B 
2 dt
∆is,C Gm Gm Gs us,C Cm Cm Cs us,C
(60)

         
∆ir,A Gs Gm Gm ur,A Cs Cm Cm ur,A
1 d
 ∆ir,B  =  Gm Gs Gm  ·  ur,B  +  Cm Cs Cm  ·  ur,B  (61)
2 dt
∆ir,C Gm Gm Gs ur,C Cm Cm Cs ur,C

In case of a zero sequence mutual impedance are defined (see chapter 4.1). An addition zero
sequence voltage u0 is added, if not u0 = 0.

Where
n
X d
U0 = (R0mi · I0i + L0mi · I0i ) kV
i
dt
√
U0 · 3
u0 = √ p.u.
2 · U nom

and

• i index of coupled lines

DIgSILENT PowerFactory 2024, Technical Reference 30

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

• R0mi is the zero sequence mutual resistance in Ω

• L0mi is the zero sequence mutual inductance in H

• I0i is the zero current in kA of the coupled lines
• U nom is the nominal voltage in kV of connected terminal i

The zero sequence current I0i is negated in case of the orientation is in opposite direction.

7.2 Distributed Parameter Model

The distributed parameter model cannot be directly used for EMT simulations because the el-
ements of the equivalent circuit are a function of the frequency, as shown in equations (39).
To make the model usable for EMT simulations, further assumptions have to be made. The
resulting distributed parameters models are described below.

The EMT models of distributed parameter lines are based on Bergeron’s method for solution in
the time domain. The following options are supported:

• Constant parameter model

• Frequency dependent parameter (modal domain)

• Frequency dependent parameter (phase domain)

These options and their associated settings can be found on the EMT-Simulation page of the
line element (ElmLne) and the line coupling (ElmTow) element. The frequency dependent pa-
rameter (phase domain) model is available for three-phase, single-phase, AC/DC and DC lines.

7.2.1 Bergeron’s Method for Solution in the Time Domain

Considering the border conditions depicted in Figure 2.13, (35) can be written as:

   
Us − Is · ZC Us − Is · ZC
Ur = ·e γ·l
+ · e−γ·l (62)
2 2

   
Us − Is · ZC Us + Is · ZC
ZC · Ir = ·e γ·l
− · e−γ·l (63)
2 2

and subtracting 63 from 62:

Ur − ZC · Ir = (Us + ZC · Is ) · e−γ·l (64)

or rewritten as  
Ur Us
Ir = − Is + · e−γ·l (65)
ZC ZC

The expression U + Zc · I of the border condition at the sending end, s, is the same at the
receiving end, r, after multiplication with the propagation factor e−γ·l .

DIgSILENT PowerFactory 2024, Technical Reference 31

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

By repeating this procedure and setting the initial conditions at node r, and then travelling with
the wave from node r to node s, we obtain:

Us − ZC · Is = (Ur + ZC · Ir ) · e−γ·l (66)

or rewritten as  
Us Ur
Is = − Ir + · e−γ·l (67)
ZC ZC

Equations (64) and (66), or equations (65) and (67), define Bergeron’s equations in the fre-
quency domain. The method models the line using controlled current sources, J, with parallel
admittances, YC , at both ends as shown in Figure 7.1 or alternatively, using a controlled voltage
source, V, in series with the impedance, ZC , as shown in Figure 7.2, where:

 
Ur
Jr = Ir + · e−γl (68)
ZC

Vr = (Ur + Ir · ZC ) · e−γl (69)

Figure 7.1: Bergeron’s method: equivalent circuit with controlled current sources

Figure 7.2: Bergeron’s method: equivalent circuit with controlled voltage sources

The inverse Fourier transform is applied to transform the set of equations into the time domain:

us (t) = F −1 ZC · Is + (Ur + ZC · Ir ) · e−γ·l


(70)

DIgSILENT PowerFactory 2024, Technical Reference 32

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

ur (t) = F −1 ZC · Ir + (Us + ZC · Is ) · e−γ·l


(71)

or rewritten:    
−1 Us Ur −γ·l
is (t) = F − Ir + ·e (72)
Zc Zc

   
−1 Ur Us −γ·l
ir (t) = F − Is + ·e (73)
Zc Zc

where both the characteristic impedance:

s
R0 + jω · L0
ZC = ZC (ω) = (74)
G0 + jω · C 0

and the propagation constant:

p
γ = γ(ω) = α(ω) + jβ(ω) = (R0 + jω · L0 ) · (G0 + jω · C 0 ) (75)

are frequency dependent, even for constant per unit length line parameters R’, L’, G’ and C’.

7.2.2 Constant Parameter Model

The constant distributed parameter model in PowerFactory is based on Bergeron’s method,

which calculates the voltages and currents at one end of the line based on the voltage and
current at the other end delayed in time (i.e. the history term). Specifically, PowerFactory
uses the lossy constant distributed parameter model described in [3]. This model is valid for
lossy lines with a series resistance, R0 , and a negligible shunt conductance (a defined shunt
conductance is considered internally as lumped at each terminal but not explicitely as part of
the constant parameter model described below). These kinds of lines can then be modelled
as single or multiple sections of lossless lines with resistances lumped in three places: R/4 at
both ends of the line, and R/2 in the middle [3]. This model provides sufficient accuracy only
if R/4 << Z, and hence should not be used for lines with very high resistance. PowerFactory
emits an error when R/4 > Z, and a warning when R/4 > 0.05 · Z.

The surge impedance is no longer calculated according to (74), but instead as:

r
L0
ZC = (76)
C0
being real and constant. The damping coefficient α = 0, and hence from (75):
√
γ = jβ = jω L0 · C 0 (77)

The propagation velocity is the same regardless of frequency and is given by:
ω 1
v= =√ (78)
β L · C0
0

which allows the following definition of a travel time (frequency independent):

l √
τ= = l · L0 · C 0 (79)
v

DIgSILENT PowerFactory 2024, Technical Reference 33

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

In terms of travel time, the propagation constant can be rewritten as:

√
γ · l = jβ · l = jω L0 · C 0 · l = jω · τ (80)
and after simplifications we can rewrite (70) and (71) as:

Us = ZC · Is + Vr
(81)
Ur = ZC · Ir + Vs

with:
Vr = (Ur + ZC · Ir ) · e−jω·τ
(82)
Vs = (Us + ZC · Is ) · e−jω·τ

The inverse Fourier transform of the phase shift e−jω·τ in the frequency domain becomes a
delay τ in the time domain, and the set of equations (81) and (82) transforms to:

us (t) = ZC · is (t) + us (t − τ )
(83)
ur (t) = ZC · ir (t) + ur (t − τ )

where the history terms are defined as:

us (t − τ ) = ur (t − τ ) + ZC · ir (t − τ )
(84)
ur (t − τ ) = us (t − τ ) + ZC · is (t − τ )

This can be reformulated in terms of current sources:

us (t)
is (t) = ZC + is (t − τ )
(85)
ur (t)
ir (t) = ZC + ir (t − τ )

where the history terms are defined as:

is (t − τ ) = − ur Z
(t−τ )
C
− ir (t − τ )
(86)
ir (t − τ ) = − us (t−τ
ZC
)
− is (t − τ )

The lumping of resistances according to the lossy model requires the modification of the impedance
according to [3]:
R
Zmod = Z + (87)
4
where Z is calculated according to (76). This results in the following modifications to the history
terms in (86) [3]:
R/4
is (t − τ ) = − Z 2Z [ur (t − τ ) + (Z − R
4) · ir (t − τ )] − 2
Zmod
[us (t − τ ) + (Z − R
4) · is (t − τ )]
mod

R/4
ir (t − τ ) = − Z 2Z [us (t − τ ) + (Z − R
4) · is (t − τ )] − 2
Zmod
[ur (t − τ ) + (Z − R
4) · ir (t − τ )]
mod
(88)

In PowerFactory, the settings for this model can be adjusted on the EMT-Simulation page of the
line element (ElmLne) or line coupling element (ElmTow) as follows:

DIgSILENT PowerFactory 2024, Technical Reference 34

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

• Line Model: Constant parameter

• Frequency for travel time estimation: enter a representative frequency for the transient
under analysis. This frequency is used in (59) to calculate the propagation constant. In
the case of a non-transposed line, the modal transformation matrix will also be calculated
at this frequency.

Note: press the Calculate Line Parameters button any time you modify these parameters. Pow-
erFactory will then calculate the propagation factor and the surge impedance at the specified
frequency and initialise the model.

7.2.3 Frequency Dependent Parameter (Modal) Model

With the exception of lossless and distortion-less lines, the characteristic impedance, ZC , and
propagation constant, γ, are frequency dependent. The variation of Zc and γ with frequency is
most pronounced in the zero sequence mode and hence frequency-dependent models should
be used when zero sequence currents or voltages are involved (i.e. in a single phase-to-ground
fault).

Approximation by Rational Functions

To handle frequency-dependent parameters, PowerFactory uses the approach proposed by J.

Marti [3, 7]. The characteristic impedance and the propagation factor are developed in rational
functions and then the poles and zeros of the rational expressions calculated using a Bode
approximation.

For the propagation factor, A (ω) = e−γ(ω)·l :

(s + z1 ) · (s + z2 ) . . . (s + zn )
Aapp (s) = e−s·τmin · k · (89)
(s + p1 ) · (s + p2 ) . . . (s + pm )

with s = jω and n < m.

Similarly for the characteristic impedance, ZC :

(s + z1 ) · (s + z2 ) . . . (s + zn )
Zc−app (s) = k · (90)
(s + p1 ) · (s + p2 ) . . . (s + pn )

with s = jω. This expression can then be expanded into partial fractions:

k1 k2 kn
Zc−app (s) = k0 + + ··· (91)
(s + p1 ) (s + p2 ) (s + pn )

The accuracy of the model depends on the quality of the rational function approximations for A
and Zc. To verify the approximation PowerFactory provides plots the exact and approximated
solutions of A and Zc on the EMT-Simulation page (Mode tabs) of the line (ElmLne) and line
coupling (ElmTow) elements as shown in Figure 7.3. Note that the plots for each mode are
displayed on their own tabbed pages. Right click on the plot to zoom in and out.

DIgSILENT PowerFactory 2024, Technical Reference 35

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
The accuracy of the model depends on the quality of the rational function approximations for A and
Zc . To verify the approximation PowerFactory plots the exact and approximated solutions of A and
Zc in the EMT-Simulation tab page of the line (ElmLne) and line coupling (ElmTow) elements as shown
in Figure 14. Note
7 EMTthat you will need to jump between dialog pages using
Simulation to display the plots.
Right click on the plot to zoom in and out.

Figure 7.3: Bode approximations of A and Zc for the zero-sequence

TechRef Overhead Lines Models

Solution in the Time Domain

In this section, only equations for the equivalent circuit with current sources are described.
Similar equations are used for the equivalent circuit with voltage sources.

Explicitly writing the frequency-dependent parameters, the input current at node s is:

Us
Is = + Jr (ω)
ZC (ω)

 
Ur
Jr (ω) = − Ir + · A (ω)
ZC

The inverse Fourier transform of the controlled current source, Jr , can be evaluated by means
of the convolution integral and hence:

∞  
ir (t − u)
Z
jr (t) = − ir (t − u) + · a (u) · du (92)
0 Zc

with a (t) = F −1 {A (ω)}, τ min is the travel time of the fastest waves and τ max is the travel time
of the slowest waves. The convolution integral then only needs to be evaluated between τ min
and τ max because a (t) is zero up to t = τ min and tends to zero for t → τ max.

a(t) is the inverse Fourier transform of A(ω). After A(ω) is expanded into partial fractions, the
inverse Fourier transform then becomes a sum of exponentials:


0 for t < τmin
aapp (t) = (93)
k1 e−p1 (t−τmin ) + k2 e−p2 (t−τmin ) · · · km e−pm (t−τmin ) for t ≥ τmin

Similarly, the inverse Fourier transform of (91) results in exponential terms of the form e−t/RC
that corresponds to an RC network as shown in Figure 7.4 where:

DIgSILENT PowerFactory 2024, Technical Reference 36

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

ki 1
R0 = k0 and Ri = , Ci = with i = 1 . . . n (94)
pi ki

Figure 7.4: Equivalent circuit for the frequency-dependent parameter model

Then with a(t) being a sum of exponential functions and Zc developed as an RC-network, equa-
tion (92) can be solved using recursive convolution.

Data for the distributed frequency-dependent parameters model can be adjusted on the EMT-
Simulation page of the line element (ElmLne) or line coupling element (ElmTow) as follows:

• Line Model: Frequency dependent parameter

• Frequency for travel time estimation: enter a representative frequency for the transient
under analysis. The modal transformation matrix is calculated at this frequency, and after-
wards remains real and constant.
• Min and Max. Frequency of parameter fitting: enter the minimum and maximum frequency
for the approximation by rational functions of the propagation factor (82) and the charac-
teristic impedance (85).

• Tolerance for Bode approximation: the maximum allowable error in % for the Bode ap-
proximation of the propagation factor (82) and the characteristic impedance (85). The
lower the tolerance the higher the number poles and zeros of the approximated rational
expressions.

Note: press the Calculate Line Parameters button any time you modify these parameters or
enter new ones. PowerFactory will then calculate the propagation factor and the characteristic
impedance at the specified frequency and set up the model.

7.2.4 Frequency Dependent Phase-Domain Models

In general, transmission lines can be characterised by two frequency-dependent matrix transfer

functions:

• The propagation function, A(ω)

DIgSILENT PowerFactory 2024, Technical Reference 37

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

• The characteristic admittance, Yc (ω) (or characteristic impedance Zc (ω))

These are expressed mathematically as:

√
A = e− YZ·l
(95)

and
√
Yc = Z−1 · ZY (96)

where Z and Y are the impedance and admittance matrices, respectively, and l is the line
length. The time-domain simulation could be formulated using the inverse Fourier transform
of the above functions and then solving the associated equations in the time domain using
numerical convolution. However, the preferred time-domain solution uses rational function ap-
proximations of low order to ensure a computationally-efficient solution. In PowerFactory, these
rational function approximations are obtained using vector fitting [5].

7.2.4.1 Rational Function Approximation

The accuracy of the rational function approximations of A and Yc (or Zc ) strongly influences the
quality of the time-domain solution. The fitting of A and Yc (or Zc ) in the phase domain means
that the phase-domain transfer functions will be intrinsically stable [4].

7.2.4.2 Approximation of Yc or Zc

The characteristic admittance, Yc , (or characteristic impedance, Zc ), are smooth functions of

frequency, hence the fitting is straightforward and can be carried out directly in the phase domain
[2]. The trace of Yc is fitted in order to calculate the poles, and the residues are subsequently
calculated per element (k,l) of Yc using the following approximation:
N
X ri,m
(Yc (s))k,l = d + (97)
m=1
s − pi,m

where constant d is real, N is the number of poles, ri,m are the residues, and pi,m are the poles.
The residues and poles may be real or in complex conjugate pairs. Fitting is performed in the
phase domain using vector fitting, and all elements of Yc get identical poles. This procedure is
used for the fitting of Yc (or Zc ) for both the ULM and FDM.

7.2.4.3 Universal Line Model

The Universal Line Model (ULM) proposed in [2] offers high accuracy and a phase-domain
formulation.

Approximation of A The fitting of the propagation function, A, is difficult because it contains

modal components which have differing time delays [2]. In practicality, the differences in the
time delays can be significant in the case of cables. This is due to the different permittivities in
the insulation. Hence, the approximation of A is performed using a two-step approach; (i) fitting
in the modal domain; followed by (ii) final fitting in the phase domain.

Modal-Domain Fitting of A: The modes of A are calculated using a frequency-dependent

transformation matrix, T. The modal-domain fitting is comprised of three procedures:

DIgSILENT PowerFactory 2024, Technical Reference 38

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

Backwinding: Multiplication with a factor, e(j·ω·τ ) , in order to remove most of the oscillatory
behaviour of the elements of A. These elements are oscillating functions of frequency due to
the time delay of the line [4]. The diagonal elements of the modal propagation matrix, Am , can
be expressed as [8]:
ω
−αi (ω)+j· v ·l
am
i (ω) = e
i (ω) (98)
where α is the attenuation, v is the velocity, j is the imaginary unit and l is the line length. Each
mode i is approximated by:
N
X ri,m
am
i (s) ≈ · e−s·τi (99)
m=1
s − pi,m

where N is the number of poles for mode i, ri,m are the residues, and pi,m are the poles. The
residues and poles may be real or in complex conjugate pairs.

Collapsing: The process of replacing modes having almost equal time delays with a single mode
equal to the average of the modes. The criterion used for the formation of groups is provided in
[2]. The concept of “groups” thereby replaces that of modes.

Fitting: Vector fitting is used to obtain the rational function approximation and all poles contribute
to all elements of A.

Phase-Domain Fitting of A: The final fitting of A (i.e. the calculation of residues) is done in
the phase domain using the poles and time delays found via fitting in the modal domain. Each
element has the form:
n XN
X ri,m
A(s) ≈ ( ) · e−s·τi (100)
i=1 m=1
s − pi,m

where n is now the number of groups, and the poles, pi,m , and the time delays, τi are known.

7.2.4.4 Frequency Dependent Model (FDM)

The Frequency Dependent Model (FDM) introduced in [6] offers improved accuracy and stability
via the simultaneous fitting of poles and residues in the phase domain. High residue-pole ratios
are thereby avoided, making the FDM less prone to divergence due to such high ratios in EMT
simulations.

Approximation of A The difficulties associated with the fitting of the propagation function, A,
were mentioned in Section 7.2.4.3. The approximation of A for the FDM is performed by fitting
the modal contributions in the phase domain using a common set of poles for each modal
contribution, i, below. The poles and residues of A are obtained simultaneously, as follows:
n
A(s) ∼
X
= Âi · e−s·τi (101)
i=1

where n is the number of modal contribution groups. Fitting is then performed with the constant
time delay, τi , removed:
Mi
X Ri,j
Âi (s) ≈ (102)
j=1
s − pi,j

where Mi is the order of the approximation for the ith modal propagation function.

Fitting Procedure The following steps are repeated, increasing the number of poles until either:
(i) the user-defined RMS error threshold has been met; or (ii) the user-defined maximum number
of poles has been reached:

DIgSILENT PowerFactory 2024, Technical Reference 39

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

1. Vector fitting is applied iteratively to reduce the RMS error. Iteration stops if either the
user-defined maximum RMS error or the user-defined maximum number of iterations is
reached.
2. If using option Increase and flip unstable poles, unstable poles are flipped to the left-half
plane and the RMS error is recalculated based on the resulting set of poles. This generally
worsens the RMS error, as expected.

3. If using option Increase until rms error is met, the RMS error is calculated based on the
unadjusted set of poles. This generally results in a lower RMS error than the previous
option, however sometimes at the cost of unstable poles which can result in time-domain
divergence in some cases.
4. For both of these options, if the RMS error does not meet the RMS error threshold, the
number of poles is increased, and the algorithm continues from the first step.

Following the fitting process, PowerFactory reports the RMS error and the poles obtained by
vector fitting (if option Output poles is selected).

The stability of the time-domain solution is dependent not only upon the accuracy of the rational
function approximation (i.e. the resulting RMS error), but also upon the proximity of the poles to
each other, and the ratio between residue and pole.

For Yc and Zc , a worse RMS error is usually obtained when fitting starting at very low frequen-
cies, and a better RMS error is usually obtained using a higher-order approximation (i.e. more
poles).

For A, a low number of poles is usually sufficient for accurate time-domain simulation, which as-
sists in avoiding high residue-to-pole ratios that tend to occur with higher-order approximations
(in particular for the ULM).

7.2.4.5 Options

The following inputs are available for vector fitting:

• Min. Frequency of Parameter Fitting: Lower bound of the frequency range for the fitting
in Hz

• Max. Frequency of Parameter Fitting: Upper bound of the frequency range for the fitting
in Hz
• RMS error: The target RMS error to be obtained by the vector fitting.
• Max. number of iterations: The maximum number of iterations used when performing
the vector fitting.
• Max. number of poles: The maximum number of poles allowed. A larger number of
poles may increase accuracy but will cause a subsequently slower performance.
• Number of poles: The number of poles found will be determined by which option is
selected:
– Increase and flip unstable poles: The number of poles will be increased and if any
pole resides in the right half (unstable region) of the complex plane it will be flipped
into the left-half plane.

DIgSILENT PowerFactory 2024, Technical Reference 40

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

– Increase until rms met: The number of poles will be increased until the target rms
error is met. This may yield a lower rms error than using Increase and flip unstable
poles at the risk of finding poles in the right-half plane (i.e. unstable poles). These
poles will likely cause time-domain divergence but can be seen if option Output poles
(described below) is selected. Hence, this option must be used with caution.
• Stability margin: The margin from the vertical axis at which poles will be placed. For
example, an entry of -1 will allow any negative pole to be placed no closer than this to the
vertical (imaginary) axis in the left-half plane.
• Fine-tune delays: Applies an optimisation procedure to improve the vector fitting by read-
justing the time delays together with the poles and residues.
• Apply DC correction: Improves the fit near DC via the use of a functional reformulation.
• Output poles: The poles will be printed in the output window.

7.3 Diagonalisation

The modal-domain models presented in §7.2 implicitly assume a single-phase line. In real-
ity however, distributed parameter models are required for three-phase long lines or for trans-
mission systems with multiple 3-phase circuits. To handle this, equations (59) - (61) have to
be diagonalised. After diagonalisation, the mutually-coupled equations of the 3-phase system
transform to 3 independent, and hence decoupled, single-phase systems.

The diagonalisation in PowerFactory is carried out as follows:

• Balanced lines: the impedance and admittance matrices of these lines are diagonal-cyclic,
i.e. Z/Y-matrices of the form (8). This is normally the case for transposed lines. To diago-
nalise the matrices, PowerFactory uses the transformation into symmetrical components
according to (19). The transformation matrix is known a-priori and is constant (not fre-
quency dependent).
• Unbalanced lines: this is typically the case for untransposed lines. The Z/Y-matrices are
no longer diagonal-cyclic; hence to diagonalise them, a transformation into modal compo-
nents is required. In this case, the transformation matrices are not known a-priori but are
determined from an eigenvalue and eigenvector calculation. In this case, the transforma-
tion matrices are frequency dependent.

For unbalanced lines and steady-state calculations (i.e. harmonic load flow or frequency sweep),
PowerFactory calculates the transformation matrices, and therefore the eigenvalues and eigen-
vectors, at each frequency of interest to account for the frequency dependency of the transfor-
mation matrices.

In the EMT-Simulation, the transformation matrix is calculated at a single frequency, i.e. at

the one specified by the user on the EMT-Simulation page of the ElmLne or ElmTow element
dialog. The transformation matrix then remains constant. Furthermore the approximation of the
complex transformation matrix (eigenvectors) is made only using the real part.

7.4 Consideration of Line Shunt Compensation

The following two configurations are considered:

1. The line compensation is a shunt reactor (i.e. when Q1 > 0)

DIgSILENT PowerFactory 2024, Technical Reference 41

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
7 EMT Simulation

2. The line compensation is a shunt capacitor

7.4.0.6 Shunt Reactor

The current through the reactor is first transformed to the αβ0 system, and the voltage across
the reactor is then calculated as follows:
diLα
uLα = <(z 1comp ) · iα + =(z 1comp )/ω0 ·
dt
diLβ
uLβ = <(z 1comp ) · iβ + =(z 1comp )/ω0 · (103)
dt
diL0
uL0 = <(z 0comp ) · i0 + =(z 0comp )/ω0 ·
dt

The α, β, 0 voltages are then transformed back to phase quantities (uLa , uLb , uLc ) to formulate
the differential equations using ua , ub and uc (the terminal voltages of the line end to which the
reactor is connected). The corresponding impedances are obtained from the admittances:
1
z 1comp =
y1
comp
(104)
1
z 0comp =
y0
comp

7.4.0.7 Shunt Capacitor

The currents though the capacitor are:

duα
iCα = <(y 1 ) · uα + =(y 1 )/ω0 ·
comp comp dt
duβ
iCβ = <(y 1 ) · uβ + =(y 1 )/ω0 · (105)
comp comp dt
du0
iC0 = <(y 0 ) · u0 + =(y 0 )/ω0 ·
comp comp dt
The voltages α, β, 0 are calculated from the terminal phase voltages (ua , ub , uc ) using a stan-
dard transformation. The currents through the capacitor are then transformed back to phase
quantities and subtracted from the line currents, ia,b,c .

DIgSILENT PowerFactory 2024, Technical Reference 42

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
A Parameter Definitions

A Parameter Definitions

Table A.1: Input parameter of the line type (TypLne)

Name Description Unit Range Default Symbol

loc name Name
uline Rated Voltage kV x>=0 0
sline Rated Current kA x>0 1
InomAir Rated Current (in air) kA x>0 1
frnom Nominal Frequency Hz x>=0 50
aohl Cable / OHL (overhead line) cab
systp System Type AC:DC AC
nlnph No. of Phases 01:02:03 3
nneutral No. of Neutrals 00:01 0
0
rline Parameters per Length 1,2- Ohm/km x>=0 0 R1
Sequence: Resistance R’
(20◦ C)
0
xline Parameters per Length 1,2- Ohm/km 0 X1
Sequence: Reactance X’
0
lline Parameters per Length 1,2- mH/km x>=0 0 L1
Sequence: Inductance L’
0
rline0 Parameters per Length Zero Ohm/km x>=0 0 R0
Sequence: Resistance R0’
0
xline0 Parameters per Length Zero Ohm/km x>=0 0 X0
Sequence: Reactance X0’
0
lline0 Parameters per Length Zero mH/km x>=0 0 L0
Sequence: Inductance L0’
0
rnline Parameters per Length, Neu- Ohm/km x>=0 0 Rn
tral:
Resistance Rn’
0
xnline Parameters per Length, Neu- Ohm/km x>=0 0 Xn
tral:
Reactance Xn’
0
lnline Parameters per Length, Neu- mH/km x>=0 0 Ln
tral:
Inductance Ln’
0
rpnline Parameters per Length, Ohm/km x>=0 0 Rpn
Phase-Neutral Coupling:
Resistance Rpn’
0
xpnline Parameters per Length, Ohm/km x>=0 0 Xpn
Phase-Neutral Coupling:
Reactance Xpn’
0
lpnline Parameters per Length, mH/km x>=0 0 Lpn
Phase-Neutral Coupling:
Inductance Lpn’
◦
tmax Parameters per Length 1,2- C x>=20 80
Sequence: Max. Operational
Temperature
rline tmax Parameters per Length 1,2- Ohm/km x>=0 0
Sequence: Resistance R’ at
max. operational tempera-
ture

DIgSILENT PowerFactory 2024, Technical Reference 43

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
A Parameter Definitions

alpha Parameters per Length 1,2- 1/K x>=0 0.00403 α

Sequence: Temperature Co-
efficient
mlei Parameters per Length 1,2- Al
Sequence: Conductor Mate-
rial
0
bline Parameters per Length 1,2- uS/km 0 B1
Sequence: Susceptance B’
0
cline Parameters per Length 1,2- uF/km 0 C1
Sequence: Capacitance C’
tline Parameters per Length 1,2- 0 tgδ1
Sequence: Ins. Factor
0
gline Parameters per Length 1,2- uS/km 0 G1
Sequence: Conductance G’
0
bline0 Parameters per Length Zero uS/km 0 B0
Sequence: Susceptance B0’
0
cline0 Parameters per Length Zero uF/km 0 C0
Sequence: Capacitance C0’
Ices Parameters per Length Zero A/km 0
Sequence: Earth-Fault Cur-
rent
tline0 Parameters per Length Zero 0 tgδ0
Sequence: Ins. Factor
0
gline0 Parameters per Length Zero uS/km 0 G0
Sequence: Conductance G0’
0
bnline Parameters per Length, Neu- uS/km x>=0 0 Bn
tral: Susceptance Bn’
0
cnline Parameters per Length, Neu- uF/km x>=0 0 Cn
tral: Capacitance Cn’
0
bpnline Parameters per Length, uS/km x>=0 0 Bpn
Phase-Neutral Coupling:
Susceptance Bpn’
0
cpnline Parameters per Length, uF/km x>=0 0 Cpn
Phase-Neutral Coupling:
Capacitance Cpn’
◦
rtemp Max. End Temperature C x>0 80
Ithr Rated Short-Time (1s) Cur- kA x>=0 0
rent (Conductor)
picln Inrush Peak Current: Ratio p.u. 0
Ip/In
pitln Inrush Peak Current: Maxi- s 0
mum Time
fcharC1 Frequency Depen-
dency of Pos.-Sequence
Capacitance: C1’(f)
(ChaPol,ChaVec,ChaMat)
fcharC0 Frequency Depen-
dency of Zero-Sequence
Capacitance: C0’(f)
(ChaPol,ChaVec,ChaMat)
fcharR1 Frequency Dependen-
cies of Pos.-Sequence
Impedance: R1’(f)
(ChaPol,ChaVec,ChaMat)

DIgSILENT PowerFactory 2024, Technical Reference 44

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
A Parameter Definitions

fcharL1 Frequency Dependen-

cies of Pos.-Sequence
Impedance: L1’(f)
(ChaPol,ChaVec,ChaMat)
fcharR0 Frequency Dependen-
cies of Zero-Sequence
Impedance: R0’(f)
(ChaPol,ChaVec,ChaMat)
fcharL0 Frequency Dependen-
cies of Zero-Sequence
Impedance: L0’(f)
(ChaPol,ChaVec,ChaMat)
pStoch Stochastic model (StoTyplne)
manuf Manufacturer
chr name Characteristic Name
for name Foreign Key
dat src Data source MAN
doc id Additional Data ()
desc Description
miso Insulation Material
iopt cnd Cable is mlt
cmeth Installation Method (IEC 364) C
iopt ord Conductors tre
qurs Nominal Cross Section mm*2 0
cabdiam Outer Diameter mm 0
iopt dir Arrangement hor
lcost Line Cost $/km x>=0 0

DIgSILENT PowerFactory 2024, Technical Reference 45

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
B References

B References

[1] Technical Reference Overhead Lines Constants, 2009.

[2] M. Tartibi A. Morched, B. Gustavsen. A universal model for accurate calculation of elec-
tromagnetic transients on overhead lines and underground cables. IEEE Transactions on
Power Delivery, 14, 1999.

[3] H. Dommel. EMTP Theory Book. Microtran Power System Analysis Corporation, 1 edition,
1996.
[4] B. Gustavsen and A. Semlyen. Combined phase and modal domain calculation of trans-
mission line transients based on vector fitting. IEEE Transactions on Power Delivery, 13,
1998.

[5] B. Gustavsen and A. Semlyen. Rational approximation of frequency domain responses by

vector fitting. IEEE Transactions on Power Delivery, 14(3):1052–1061, 1999.
[6] I. Kocar and J. Mahseredjian. Accurate frequency dependent cable model for electromag-
netic transients. IEEE Transactions on Power Delivery, 2015.

[7] J. R. Marti. The Problem of Frequency Dependence in Transmission Line Modelling. PhD.
Thesis. The University of British Columbia, 1981.
[8] L. De Tommasi and B. Gustavsen. Accurate transmission line modeling through optimal
time delay identification. Proceedings of the International Conference on Power Systems
Transients, IPST07, 2007.

DIgSILENT PowerFactory 2024, Technical Reference 46

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
List of Figures

List of Figures

1.1 Positive sequence input parameters . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1 Equivalent PI-circuit of the line for lumped parameters . . . . . . . . . . . . . . . 5

2.2 Lumped parameter model for a single-phase AC line . . . . . . . . . . . . . . . . 6

2.3 Equivalent circuit of the single-phase line with neutral conductor model . . . . . . 7

2.4 Equivalent circuit of the two-phase line model . . . . . . . . . . . . . . . . . . . . 8

2.5 Equivalent circuit of the two-phase line with neutral conductor model . . . . . . . 9

2.6 Equivalent circuit of the three-phase line . . . . . . . . . . . . . . . . . . . . . . . 10

2.7 Equivalent circuit for the 3-phase line with neutral conductor . . . . . . . . . . . . 11

2.8 Phase to phase measurement loop . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.9 Neutral to ground measurement loop . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.10 Phase – PE (earth wire) measurement . . . . . . . . . . . . . . . . . . . . . . . . 13

2.11 Phase – neutral measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2.12 Incremental model for a line of elemental length . . . . . . . . . . . . . . . . . . . 16

2.13 Equivalent pi-circuit for the line with distributed parameters in the frequency domain 17

4.1 Example for zero sequence mutual impedance coupling . . . . . . . . . . . . . . 25

4.2 Orientation of zero sequence mutual impedance . . . . . . . . . . . . . . . . . . 26

4.3 Short-circuit at line with zero sequence mutual impedance . . . . . . . . . . . . . 27

7.1 Bergeron’s method: equivalent circuit with controlled current sources . . . . . . . 32

7.2 Bergeron’s method: equivalent circuit with controlled voltage sources . . . . . . . 32

7.3 Bode approximations of A and Zc for the zero-sequence . . . . . . . . . . . . . . 36

7.4 Equivalent circuit for the frequency-dependent parameter model . . . . . . . . . 37

DIgSILENT PowerFactory 2024, Technical Reference 47

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)
List of Tables

List of Tables

1.1 Overview of line models available in PowerFactory . . . . . . . . . . . . . . . . . 1

A.1 Input parameter of the line type (TypLne) . . . . . . . . . . . . . . . . . . . . . . 43

DIgSILENT PowerFactory 2024, Technical Reference 48

Overhead Line Models (ElmLne, TypLne, TypGeo, TypTow)