DIgSILENT

PowerFactory 2018

Technical Reference Documentation

Cable System
ElmCabsys, TypCabsys

I N T EG R AT E D P O W E R S Y S T EM A N A LY S I S S O F T WA R E F O R
T R A N S M I S S I O N / D I S T R I BU T I O N / I N D U S T RY / G EN E R AT I O N / I N T EG R AT I O N O F R EN E WA B L E S
 Publisher:
DIgSILENT GmbH
Heinrich-Hertz-Straße 9
72810 Gomaringen / Germany
Tel.: +49 (0) 7072-9168-0
Fax: +49 (0) 7072-9168-88
[email protected]

Please visit our homepage at:

http://www.digsilent.de

Copyright © 2017 DIgSILENT GmbH

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

December 4, 2017
PowerFactory 2018
Revision 1
Contents

Contents

1 Introduction 1

2 Definition of the Cable System 2

2.1 The Single Core Cable Type (TypCab) . . . . . . . . . . . . . . . . . . . . . . . . 2

2.1.1 Filling factor of conducting layers . . . . . . . . . . . . . . . . . . . . . . . 2

2.1.2 Temperature coefficient . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.2 The Cable System Type (TypCabsys) . . . . . . . . . . . . . . . . . . . . . . . . 3

3 Calculation of Electrical Parameters 3

3.1 Internal Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

3.2 Internal Admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3.3 Semiconducting Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3.4 Parallel Single-Core Cables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3.4.1 Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3.4.2 Admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.5 Pipe Type Cable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.5.1 Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.5.2 Admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

4 EMT Simulation 12

4.1 Universal Line Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4.1.1 Rational Function Approximation . . . . . . . . . . . . . . . . . . . . . . . 12

4.1.2 Fitting Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

A Parameter Definitions 15

B Calculation Results 17

C References 20

List of Figures 21

List of Tables 22

DIgSILENT PowerFactory 2018, Technical Reference i

Cable System (ElmCabsys, TypCabsys)
1 Introduction

1 Introduction

This document describes the definition of a cable system in terms of its geometry, the proper-
ties of the conducting, semi-conducting and insulating layers, installation characteristics (buried
directly underground; in a pipe). In addition, the calculation of frequency-dependent electrical
parameters of the cable system are described.

The definition of a frequency-dependent cable system in PowerFactory with help of two type
objects: a single core cable type TypCab which describes the construction characteristics of the
cable and a cable system type TypCabsys, which defines the coupling between phases, i.e. the
coupling between the single core cables in a multiphase/multi-circuit cable system.

A built-in cable constants function in the cable system type calculates then the frequency-
dependent electrical parameters (impedance and admittance matrices). The function can han-
dle coaxial cables consisting of a core, sheath and armour directly underground or installed in
pipes (pipe-type cables). This function can be started in a stand-alone mode, via the Calcu-
late button on the edit dialogue of the cable system, in which case the results are printed to
the output window. Alternatively, the function may be automatically called by various simulation
functions in PowerFactory , i.e. when running a frequency scan or when adjusting the model for
an EMT simulation.

Finally, it should be noted that the cable system type (TypCabsys) supports the definition of the
cable in terms of geometrical data; if the cable is to be defined in terms of electrical data, the
reader is referred to [3], in which case the general line/cable element (ElmLne) in PowerFactory
should be used instead.

DIgSILENT PowerFactory 2018, Technical Reference 1

Cable System (ElmCabsys, TypCabsys)
2 Definition of the Cable System

2 Definition of the Cable System

2.1 The Single Core Cable Type (TypCab)

The single core cable type TypCab supports up to three tubular conducting layers in a coaxial
arrangement, i.e. core, sheath and armour, separated by three insulating layers. Figure 2.1
shows the typical layout of a HV AC single core cable. The model also supports the definition
of a core-outer and insulation-outer semiconducting layer.

Figure 2.1: Cross-section of a single core cable including the core, sheath and armour

Section A shows the complete list of input parameters including units, range and the symbol
used in this document. Hover the mouse pointer over the input parameters in the TypCab
dialogue to display the name of the input parameter. This is the name listed in the first column
of the table.

The input data in the TypCab dialogue is organised according to layers, i.e. the conducting,
insulation and semiconducting layers, if available. Use TypCab to enter all the geometrical data
defining the cross-section of the single core cable and the properties of all constituent materials.

2.1.1 Filling factor of conducting layers

To account for the compacting ratio of the cross-section of the conducting layers (stranded
conductors, shaped compact, etc.), the user can enter a filling factor, Cf. This filling factor is
related to the dc resistance of the cable by the following equation:

1
RDC [Ω/km] = ρ[µΩ · cm] · × 10
π · (r2 − q 2 ) · Cf

where r and q are the outer and inner radius of the conducting layer, respectively.

The user chooses the input parameter between the filling factor in % or the DC resistance in
Ohm/km by clicking the selection arrow . Note that one of them is always greyed out indicating
its dependency on the other.

DIgSILENT PowerFactory 2018, Technical Reference 2

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

2.1.2 Temperature coefficient

If the temperature dependency of line/cables option is enabled in the Load Flow calculation, the
resistivity of the conducting layers is adjusted by the following equation:

ρT = ρ20◦ C · [1 + α(T − 20)]

where α is the temperature coefficient of resistance. The resistivities and temperature coeffi-
cient of common metals are given in Table 2.1 for reference.

Table 2.1: Resistivities and temperature coefficient of resistance

Material Resistivity at 20◦ Temperature coefficient

C [µΩ · cm] at 20◦ C [1/◦ C]
Aluminium DC Current Input p.u.
Cooper, hard drawn 2.83 0,0039
Cooper, annealed 1.77 0.00382
Brass 6.4 - 8.4 0.0020
Iron 10 0.0050
Silver 1.59 0.0038
Steel 12 - 88 0.001 - 0.005

2.2 The Cable System Type (TypCabsys)

The cable system type TypCabsys is used to complete the definition of a cable system. It
defines the coupling between phases, i.e. the coupling between the single core cables in a
multiphase/multi-circuit cable system. Cables are generally laid close together so this coupling
should be taken into account.

Among other factors, this coupling depends on how the cables are laid. The PowerFactory
model supports following two options:

• Parallel single-core cables: the cables are grounded direct into ground. This is the
normal case for underground HV AC cables.
• Pipe-type cables: the cables are drawn into a pipe, usually made of steel, and the pipe
laid into ground. This is in widespread use in submarine cables.

The input parameter Buried: Direct in ground/in Pipe lets the user choose between both models.
In case of pipe-type cables additional data is required for the pipe. The complete list of input
parameters is shown in Table A.2 in Section A.

The cable system type also defines the bonding conditions of the sheath and armours when
available.

3 Calculation of Electrical Parameters

The calculation of the impedance and admittance of the cable is based on the cable constants
equations formulated by [2], and makes the following assumptions:

DIgSILENT PowerFactory 2018, Technical Reference 3

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

• Coaxial arrangement of the conducting and insulating layers inside the single core cable;
• Single core cables inside the pipe are concentric with respect to the pipe;
• Each conducting layer of the cable has constant permeability. In addition, conducting
layers are non-magnetic so that the cable model does not account for current-dependent
saturation effects;
• Displacement currents and dielectric losses of the insulating layers are negligible.

A general formulation of the series impedance and shunt admittance of the cable is given by:

∂
[U ] = −[Z] · [I] (1)
∂x

∂
[I] = −[Y ] · [U ] (2)
∂x

where [U ] and [I] are the voltage and current vectors at a distance x along the cable.

The dimension of [Z] and [Y ] depends on the total number of cables in the system and the total
number of layers per single core cable. For instance, in a three-phase cable system with three
conducting layers per single core cable (core, sheath and armour) the dimension of the [Z] (i.e.
[Y ]) results in 9 (=3 phases x 1 single core cable/phase x 3 conducting layers/cable).

[Z] and [Y ] are symmetric square matrices that can be expressed as follows:

[Z] = [Z I ] + [Z P ] + [Z C ] + [Z 0 ] (3)

[Y ] = s · [P ]−1
[P ] = [PI ] + [PP ] + [PC ] + [P0 ] (4)

where [P ] is a potential coefficient matrix and the Laplace’s operator (complex frequency).

The matrices with subscript I account for the internal impedance and admittance, respectively,
and matrices with subscript O account for the earth or air return path. In case of a pipe enclosure
cable the matrices with subscript C and P define the impedance and admittance of the pipe;
these matrices become zero if the cable is laid directly underground. In the next subsections
we will discuss the physical meaning of these sub-matrices and the formulae used to calculate
them.

The following naming convention is used throughout this document:

• Subscript I accounts for the internal impedance; subscript O for the earth or air return
path, and subscripts C and P for the pipe enclosure (if available);
• Subscripts c,s and a (lower case) are used for core, sheath and armour in the cable layer
equations;
• Subscripts i,j and k refer to the cables in the system (typically three cables in a three-
phase cable system).

DIgSILENT PowerFactory 2018, Technical Reference 4

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

3.1 Internal Impedance

The internal impedance is associated with the longitudinal voltage drop due to the magnetic
field inside the single core cable and it is given by the following equation:

       
U Z Z cs Z ca Ic Ic
∂  c   cc
U s = Z sc Z ss Z sa  · I s  = −[ZI ] · I s  (5)
∂x
Ua Z ac Z as Z aa Ia Ia

where the layer internal impedances in 5 are defined in terms of coaxial loop impedances as
follows:

Z cc = Z 11 + 2 · Z 12 + Z 22 + 2 · Z 23 + Z 33
Z cs = Z sc = Z 12 + Z 22 + 2 · Z 23 + Z 33
Z ca = Z ac = Z sa = Z as = Z 23 + Z 33
Z ss = Z 22 + 2 · Z 23 + Z 33
Z aa = Z 33 (6)

The impedances with subscripts 1, 2 and 3 are referred to as loop impedances. For instance,
Z 11 is the impedance of the innermost loop of the concentric tubular conductors and therefore
that of the loop core-sheath.

Z 11 = Z c,OU T + Z c/s,IN S + Z s,IN

Z 22 = Z s,OU T + Z s/a,IN S + Z a,IN
Z 33 = Z a,OU T + Z a,IN S
Z 12 = Z 21 = ·Z s,M U T U AL
Z 23 = Z 32 = ·Z a,M U T U AL (7)

The impedances of the tubular conductors are found with the modified Bessel functions, with
tube = c, s and a respectively:

ρm
Z tube,IN = {I0 (mq) · K1 (mr) + K0 (mq) · I1 (mr)}
2πqrD
ρm
Z tube,OU T = {I0 (mr) · K1 (mq) + K0 (mr) · I1 (mq)}
2πqrD
ρ
Z tube,M U T U AL = (8)
2πqrD

where

D = I1 (mr) · K1 (mq) − I1 (mq) · K1 (mr) (9)

s
jωµ 1
m= = (10)
ρ p

DIgSILENT PowerFactory 2018, Technical Reference 5

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

The parameter m is the reciprocal of the depth of penetration, p, and are both frequency-
dependent complex values.

Z IN S accounts for the longitudinal voltage drop due to the magnetic field in the insulating layers.
For the general case of non-concentric tubular conductors it results in:

( "  2 #)
µ0 qk di
Z IN S = jω · ln 1− (11)
2π ri qk

and in the case of concentric tubular conductors di = 0 and 11 reduces to:

 
µ0 qk
Z IN S = jω · ln
2π ri

3.2 Internal Admittance

The internal admittance matrix is associated with the capacitive coupling and dielectric losses
due to the insulating layers within the single core cable. The capacitance and dielectric losses
of each insulating layer are given by:

2πε0 εr 1
Ci = =
ln(r/q) Pi
Gi = ωCi · tg(δ) (12)

where Pi is the potential coefficient of the insulating layer.

Assuming that the single core cable consists of three layers, hence the insulation between core
and sheath, sheath and armour and outermost insulating layer of the single core cable, it follows
that:

 
Pc + Ps + Pa Ps + Pa Pa
[PI ] =  Ps + Pa Ps + Pa Pa  (13)
Pa Pa Pa

1
[CI ] = (14)
[PI ]

[Y I ] = [GI ] + jω[CI ] (15)

3.3 Semiconducting Layers

The model supports the definition of a semiconducting layer on the conductor’s outer surface
and the insulation’s outer surface. These semiconducting layers mainly influence the admittance

DIgSILENT PowerFactory 2018, Technical Reference 6

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

of the insulation. Their effect on the impedance of the conductor is minor and therefore not
considered in the model.

The capacitance and conductance of the tubular semiconducting layer are given by the following
equations:

1
CSC = 2πε0 · εrSC
ln(rSC /qSC )

2π 1
GSC = ·
ρSC ln(rSC /qSC )

where rSC and qSC are the outer and inner radius of the tubular semiconducting layer respec-
tively; εrSC is the relative permittivity and ρSC is the resistivity.

Hence the equivalent admittance of the insulation under consideration of the semiconducting
layers is calculated as:

1 1 1
= +
GiIns GIL GSC

1 1 1
= +
CiIns CIL CSC

3.4 Parallel Single-Core Cables

3.4.1 Impedance

If it assumed that i, j, k are three parallel single core cables, and each consists of core, sheath
and armour. Equation (1) can then be expanded:

   
  [Z I,ii ] [0] [0] [Z 0,ii ] [Z 0,ij ] [Z 0,ik ]   
[U ] [I i ]

..   ..
 
∂  i   
[U j ] =  . [Z I,jj ] [0]  + .  × [I j ]
[Z 0,jj ] [Z 0,jk ]  (16)
  
∂x . .
[U k ] [I k ]

.. ..
 
··· [Z I,kk ] ··· [Z 0,kk ]
 

where [Z 0,s ] and [Z 0,m ] are the self- and mutual- earth-return impedance matrices of the cable
system given as:

 
Z e,s Z e,s Z e,s
[Z 0,s ] = Z e,s Z e,s Z e,s  ...s = jj,kk,ll (17)
Z e,s Z e,s Z e,s

 
Z e,m Z e,m Z e,m
[Z 0,m ] = Z e,m Z e,m Z e,m  ...m = jk,kl,lj (18)
Z e,m Z e,m Z e,m

DIgSILENT PowerFactory 2018, Technical Reference 7

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

Z e,m is the mutual earth-return impedance between two parallel cables i, j given by:

ωµ0
Z e,jk = j [K0 (m · dik ) − K0 (m · Dik )] + (Pik + jQik ) (19)
π

and Pik + jQik the terms of the Carson’s series (see [1] for further information).

Z e,s is the self earth-return impedance of the single core cable. Its value is obtained from (19)
by replacing d with R; D with 2h; and h + y with 2h.

3.4.2 Admittance

As the cable is directly laid underground and the earth surrounding the cable is assumed to be
an equipotential surface, there is no capacitive coupling effect among the single core cables. It
then follows that [P 0 ] = Θ in equation 4 and therefore the admittance matrix of the cable results
in:

 
[PI,i ] 0 0
[P ] = [PI ] =  0 [PI,j ] 0  (20)
0 0 [PI,k ]

[Y ] = [P ]−1
   
[I i ] [U i ]
∂ 
[I j ]  = −[Y ] ·  [U j ]  (21)
∂x
[I k ] [U k ]

and the submatrices in the main diagonal represented according to equation (13).

3.5 Pipe Type Cable

3.5.1 Impedance

Assuming again a system of three single core cables, i, j, k, each of them consisting of core,
sheath and armour, equation (3) can be expanded as follows for the case of a pipe type cable:

DIgSILENT PowerFactory 2018, Technical Reference 8

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

 
   [Z I,ii ] [0] [0] 0
[U i ]


  . 
  ..

∂  [U j ]  =  [Z I,jj ] [0] 0
∂x [U k ]   ..
 
 . ··· [Z I,kk ] 0
 
[U p ]


0 0 0 0

 
[Z P,ii ] [Z P,ij ] [Z P,ik ] 0
 .. 
 . [Z P,jj ] [Z P,jk ] 0
+ 
 ...
 
··· [Z P,kk ] 0
0 0 0 0
 
[Z C1 ] [Z C1 ] [Z C1 ] Z C2
 .. 
 . [Z C1 ]
[Z C1 ] Z C2 
+ 
 ...
 
··· [Z C1 ] Z C2 
Z C2 Z C2 Z C2 Z C3
 
[Z 0 ] [Z 0 ] [Z 0 ] Z 0   
 ..


 [I i ]
 ×  [I j ] 
 . [Z 0 ] [Z 0 ] Z 0 
  
+ 
 .  [I k ] (22)
 .. · · · [Z 0 ] Z 0 
Ip


Z0 Z0 Z0 Z0


[Z P ] defines the self- and mutual- impedances of the pipe-return path of the single core cables.
A submatrix is given by:

 
Z P,ij Z P,ij Z P,ij
[Z P,ij ] = Z P,ij Z P,ij Z P,ij  (23)
Z P,ij Z P,ij Z P,ij

The self impedance with pipe-return path for the i-th cable (i=j) is:

∞
"  2n #
µ0 µr K0 (mq) X di 2µr Kn (mq)
Z P,ii = jω + (24)
2π mqK1 (mq) n=1 q nµr Kn (mq) − mqKn0 (mq)

Mutual impedance between the i-th and the j-th cables with common pipe-return path (16=j):

  
µ0 q K0 (mq)
Z P,ij = jω ln √ 2 2 + µr mqK 1 (mq)
2π di +dj −2di dj cos ϑij
∞ 
X di dj  n  )
2µr Kn (mq) 1
+ cos (nϑij ) − (25)
n=1
q2 nµr Kn (mq) − mqKn0 (mq) n

[Z C ] is the connection impedance matrix between the pipe inner and outer surfaces. The sub-
matrix [Z C1 ], [Z C2 ] and [Z C3 ] are given by:

 
Z C1 Z C1 Z C1
[Z C1 ] = Z C1 Z C1 Z C1  (26)
Z C1 Z C1 Z C1

DIgSILENT PowerFactory 2018, Technical Reference 9

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

where Z C1 , Z C2 and Z C3 are calculated using equations (8) to (11) for the impedance of tubular
conductors and tube being the pipe as follows:

Z C1 = Z pipe,OU T + Z pipe,IN S − 2 · Z pipe,M U T U AL

Z C2 = Z pipe,OU T + Z pipe,IN S − Z pipe,M U T U AL
Z C3 = Z pipe,OU T + Z pipe,IN S (27)

Finally, [Z 0 ] represent the impedance of the earth return-path of the pipe. The diagonal subma-
trix [Z 0 ] is given by:

 
Z0 Z0 Z0
[Z 0 ] = Z 0 Z0 Z 0 (28)
Z0 Z0 Z0

where Z 0 is the self earth return impedance of the pipe according to equation (19).

3.5.2 Admittance

The admittance follows the general definition in terms of the potential coefficient matrix as fol-
lows:

 

 [PI,ii ] [0] [0] 0

 . 
 ..

[PI,jj ] [0] 0
[P ] = 
 
 ..

 . ··· [PI,kk ] 0

 



0 0 0 0

 
[PP,ii ] [PP,ij ] [PP,ik ] 0
 .. 
 . [PP,jj ] [PP,jk ] 0
+ 
 ...
 
··· [PP,kk ] 0
0 0 0 0
 
[PC ] [PC ] [PC ] PC 
 ..


 . [PC ] [PC ] PC   
+ 
 . (29)
 ..

· · · [PC ] PC  

PC PC PC PC


where

 
[GI,i ] 0 0 0
 0 [GI,j ] 0 0
[Y ] = 
  + jω ([P ])−1 (30)
0 0 [GI,k ] 0
0 0 0 0

DIgSILENT PowerFactory 2018, Technical Reference 10

Cable System (ElmCabsys, TypCabsys)
3 Calculation of Electrical Parameters

   
[I i ] [U i ]
∂  [I j ]  = −[Y ] ·  [U j ] 
  
(31)
∂x [I k ] [U k ]
Ip Up

Note that equation (30) does not consider dielectric losses of the pipe.

Each of the [PI,ii ] submatrices of [PI ] is the internal potential coefficient matrix of the single
core cable according to (20).

[PP ] is the pipe internal potential coefficient matrix and defines the capacitive coupling between
the outermost layer of the single core cables and the pipe and hence the dielectric medium
between the cables and the pipe. Each of the submatrices [PP,ij ] of [PP ] is a matrix with equal
elements given in the following form:

 
PP,ij PP,ij PP,ij
[PP,ij ] = PP,ij PP,ij PP,ij  (32)
PP,ij PP,ij PP,ij

with

( "  2 #)
1 q di
Pii = ln 1− (33)
2πε0 εr Ri q

   
∞  n
1  q −
X 1 d d
i j

Pij = ln  q · · cos ϑ ij (34)
2πε0 εr  d2i + d2j − 2di dj cos ϑij n=1
n q2 

[PC ] is the potential coefficient matrix between the pipe inner and outer surfaces and hence the
capacitance due to the dielectric layer surrounding the pipe. A submatrix and the last column
and row elements are given by:

 
PC PC PC
[PC ] = PC PC PC  (35)
PC PC PC

 
1 r
PC = · ln (36)
2πε0 εr q

It is assumed in the model that the pipe is underground. Therefore the outer surface of the
insulating layer surrounding the pipe is in direct contact with the earth (equipotential surface
with U = 0). Hence no additional capacitive effect exists between the insulating layer of the pipe
and ground.

DIgSILENT PowerFactory 2018, Technical Reference 11

Cable System (ElmCabsys, TypCabsys)
4 EMT Simulation

4 EMT Simulation

Two distributed parameter models are provided in PowerFactory for the EMT simulation of cable
systems. These are the constant parameter model and the universal line model.

4.1 Universal Line Model

The Universal Line Model (ULM) proposed in [7] offers high accuracy and a phase-domain
formulation.

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

functions:

• The propagation function, A(ω)

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

These are expressed mathematically as:

√
A = e− YZ·l
(37)

and
√
Yc = Z−1 · ZY (38)

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].

4.1.1 Rational Function Approximation

The accuracy of the rational function approximations of A and Zc strongly influences the quality
of the time-domain solution. The fitting of A and Zc in the phase domain means that the phase
domain transfer functions will be intrinsically stable [6].

Approximation of Zc The characteristic impedance, Zc , is a smooth function of frequency,

hence the fitting is straightforward and can be carried out directly in the phase domain [7]. The
trace of Zc is fitted in order to calculate the poles, and the residues are subsequently calculated
per element (k,l) of Zc using the following approximation:
N
X ri,m
(Zc (s))k,l = d + (39)
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 Zc get identical poles.

DIgSILENT PowerFactory 2018, Technical Reference 12

Cable System (ElmCabsys, TypCabsys)
4 EMT Simulation

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

modal components which have differing time delays [7]. 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:

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 [6]. The diagonal elements of the modal propagation matrix, Am , can
be expressed as [4]:
ω
−αi (ω)+j· v ·l
am
i (ω) = e
i (ω) (40)

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 (41)
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
[7]. 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 (42)
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.

4.1.2 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 number of poles has been reached:

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. 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 the RMS error (after flipping) does not meet the RMS error threshold, the number of
poles is increased, and the algorithm continues from the first step.

DIgSILENT PowerFactory 2018, Technical Reference 13

Cable System (ElmCabsys, TypCabsys)
4 EMT Simulation

Following the fitting process, PowerFactory reports the RMS error and the poles obtained by
vector fitting.

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 Zc , a worse RMS error is usually obtained when fitting starting at very low frequencies, 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
assists in avoiding high residue-to-pole ratios that tend to occur with higher-order approxima-
tions.

DIgSILENT PowerFactory 2018, Technical Reference 14

Cable System (ElmCabsys, TypCabsys)
A Parameter Definitions

A Parameter Definitions

Table A.1: Input parameter of the single core cable type (TypCab)

Parameter Description Unit Range Default Symbol

loc name Name
uline Rated voltage kV x≥0 0 U
typCon Shape of the core Compact
diaCon Outer diameter of the core mm x≥0 5 r
diaTube Inner diameter of the core mm x≥0 0 q
cHasEl Exists: use this flag to en- 1
able/disable the conducting lay-
ers
rho Resistivity (20◦ C) of the conduct- µΩ · x>0 1.7241 ρ
ing layers cm
my Relative Permeability of the con- 1 µr
ducting layers
cThEl Thickness of the conducting lay- mm 2.5
ers
Cf Filling factor of the conducting % x > 0 and 100 Cf
layers x ≤ 100
rpha DC-Resistance (20◦ C) of the Ω/km 0.8780769 Rdc
conducting layers
alpha Temperature coefficient of the 1/K x≥0 0.00393 α
conducting layers
cHasIns Exists: use this flag to en- 1
able/disable the corresponding
insulation layers
tand Dielectric Loss Factor of the in- 0.02
sulating layer, i.e. tgδ of the in-
sulation. Set this value to zero to
neglect insulation losses.
epsr Relative Permittivity of the insu- 3 εr
lating layer
thIns Thickness of the insulating layer mm x≥0 1
cHasSc Exists: use this flag to en- 0
able/disable the semiconducting
layers
rhoSc Resistivity of the semiconducting µΩ · x>0 1000000
layer cm
mySc Relative permeability of the 1 µr,SC
semiconducting layer
epsrSc Relative permittivity of the semi- 3 εr,SC
conducting layer
thSc Thickness of the semiconducting mm x≥0 1
layer
◦
tmax Max. Operational Temperature C x≥0 80
◦
rtemp Max. End Temperature C x>0 80
Ithr Rated Short-Time (1s) Current kA x≥0 0
diaCab Overall Cable Diameter mm 15

DIgSILENT PowerFactory 2018, Technical Reference 15

Cable System (ElmCabsys, TypCabsys)
A Parameter Definitions

Table A.2: Input parameter of the cable system (TypCabsys)

Parameter Description Unit Range Default Symbol

loc name Name
frnom Nom. Frequency Hz 50
rhoEarth Resistivity of the earth return Ωm x>0 100
path
cGearth Conductivity of the earth return µS/cm 100
path = inverse of the earth resis-
tivity.
iopt bur To specify is the cable is laid gnd
direct in ground (parallel single
core cables) or in a pipe (pipe-
type cable)
nlcir Number of circuits defining the x≥1 1
cable system
pcab c Single core cable type: select TypCab
from the library the single core
cable type (TypCab) of each cir-
cuit
nphas Number of phases 3
dInom Rated current kA 1
red Reduced: assert this option to 0
automatically bond the sheaths
and armours of the cable. This
operation will reduced the Z/Y
matrices of the cable to nphas x
nphas.
bond Assert this option to cross bond 0
the sheaths
xy c Coordinate of Line Circuits: enter m 0
the coordinates of the single core
cables in the cable systems. Ca-
bles buried direct underground
have positive Y-distances with
respect to the ground surface.
For pipe type cables, X- and Y-
coordinates are referred to the
center of the pipe.
dep pipe Depth of the center of the m x≥0 0
pipe (parameter only required for
pipe-type cables).
rad pipe Outer Radius of the pipe m x>0 0.1
th pipe Thickness of the pipe mm x≥0 0
th ins Thickness of the pipe outer insu- mm x>0 1
lation
rho pipe Resistivity of the pipe µΩcm x>0 20
my pipe Relative permeability of the pipe x>0 1
epsr fil Relative permittivity of the filling x>0 1
material (insulating material be-
tween the single core cables and
the pipe)
epsr ins Relative permittivity of the pipe x>0 1
outer cover. Se

DIgSILENT PowerFactory 2018, Technical Reference 16

Cable System (ElmCabsys, TypCabsys)
B Calculation Results

B Calculation Results

The cable constants function in stand-alone mode can be started via the Calculate button on the
EMT-Simulation page of the dialogue of the cable system type TypCabsys. Then PowerFactory
prints the resulting impedance and admittance matrices to the output windows.

It follows an extract of the output window for a 132 kV, 3-phase cable system, 630 mm2 , directly
underground. The first two matrices correspond to the unreduced layer impedances and admit-
tances in phase components; cores first, followed by sheaths. Cables are in the same order as
the input. Rows follow real and imaginary part.

DIgSILENT PowerFactory 2018, Technical Reference 17

Cable System (ElmCabsys, TypCabsys)
B Calculation Results

The next two matrices are the impedances and admittances in symmetrical components in 0-
1-2 sequence. Idem before, cores come first followed by the sheaths. Cables are in the same
order as the input. Rows follow real and imaginary part.

DIgSILENT PowerFactory 2018, Technical Reference 18

Cable System (ElmCabsys, TypCabsys)
B Calculation Results

DIgSILENT PowerFactory 2018, Technical Reference 19

Cable System (ElmCabsys, TypCabsys)
C References

C References
[1] Technical Reference Overhead Lines Constants, 2009.

[2] A. Ametani. A General Formulation of Impedance and Admittance of Cables. IEEE Trans-
actions on Power Apparatus and Systems, PAS-99(3), May 1980.
[3] B. K. Bose. Modern Power Electronics and AC Drivers. Prentice Hall of India, 1 edition,
2008.

[4] Gustavsen B., De Tommasi, L.,. Accurate transmission line modeling through optimal time
delay identification. Proceedings of the International Conference on Power Systems Tran-
sients, IPST07, 2007.
[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] Semlyen A., Gustavsen, B.,. Combined phase and modal domain calculation of transmission
line transients based on vector fitting. IEEE Transactions on Power Delivery, 13, 1998.
[7] Gustavsen B., Tartibi M. Morched, A.,. A universal model for accurate calculation of elec-
tromagnetic transients on overhead lines and underground cables. IEEE Transactions on
Power Delivery, 14, 1999.

DIgSILENT PowerFactory 2018, Technical Reference 20

Cable System (ElmCabsys, TypCabsys)
List of Figures

List of Figures

2.1 Cross-section of a single core cable including the core, sheath and armour . . . 2

DIgSILENT PowerFactory 2018, Technical Reference 21

Cable System (ElmCabsys, TypCabsys)
List of Tables

List of Tables

2.1 Resistivities and temperature coefficient of resistance . . . . . . . . . . . . . . . 3

A.1 Input parameter of the single core cable type (TypCab) . . . . . . . . . . . . . . . 15

A.2 Input parameter of the cable system (TypCabsys) . . . . . . . . . . . . . . . . . . 16

DIgSILENT PowerFactory 2018, Technical Reference 22

Cable System (ElmCabsys, TypCabsys)