PowerFactory 2023

Technical Reference

External Grid
ElmXnet
F2023

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
[email protected]

Please visit our homepage at:

https://www.digsilent.de

Copyright © 2023 DIgSILENT GmbH

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

February 17, 2023

PowerFactory 2023
Revision 2
Contents

Contents

1 General Description 1

2 Load Flow Model 2

2.1 Primary Frequency Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2.2 Secondary Frequency Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.3 Operational Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.4 Unbalanced Load Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

3 Short-Circuit Models 6

3.1 Short-Circuit Positive Sequence Model . . . . . . . . . . . . . . . . . . . . . . . . 6

3.2 VDE/IEC Short-Circuit Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3.3 ANSI Short-Circuit Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3.4 Complete Short-Circuit Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3.5 Short-Circuit Zero Sequence Model . . . . . . . . . . . . . . . . . . . . . . . . . 7

3.6 Short-Circuit Negative Sequence Model . . . . . . . . . . . . . . . . . . . . . . . 8

4 Harmonics Model 9

4.1 Calculation of Network Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . 9

4.1.1 Calculation of Positive Sequence Voltage . . . . . . . . . . . . . . . . . . 9

4.1.2 Calculation of Negative Sequence Voltage . . . . . . . . . . . . . . . . . . 10

4.1.3 Calculation of Zero Sequence Voltage . . . . . . . . . . . . . . . . . . . . 11

4.1.4 Calculation of Zero Sequence Voltage Considering Neutral . . . . . . . . 11

4.2 Harmonic Voltages and Background Harmonics . . . . . . . . . . . . . . . . . . . 12

4.3 Frequency Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

5 RMS- and EMT-Simulation 14

5.1 Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5.2 Flux equations and time constants . . . . . . . . . . . . . . . . . . . . . . . . . . 15

5.3 Rotor equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

5.4 Stator equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

5.4.1 Stator equations used in the RMS-simulation . . . . . . . . . . . . . . . . 17

DIgSILENT PowerFactory 2023, Technical Reference i

External Grid (ElmXnet)
Contents

5.4.2 Stator equations used in the EMT-simulation . . . . . . . . . . . . . . . . 18

5.5 Mechanical Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

5.5.1 Rotor angle definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

5.5.2 Initialisation of the mechanical parameters . . . . . . . . . . . . . . . . . . 20

5.5.3 Excitation system interfacing . . . . . . . . . . . . . . . . . . . . . . . . . 20

5.6 Using default machine input parameters . . . . . . . . . . . . . . . . . . . . . . . 21

5.7 Inputs/Outputs/State Variables of the Dynamic Model . . . . . . . . . . . . . . . 21

5.7.1 Stability Model (RMS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

5.7.2 EMT-Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

List of Figures 23

List of Tables 24

DIgSILENT PowerFactory 2023, Technical Reference ii

External Grid (ElmXnet)
1 General Description

1 General Description

The PowerFactory External Grid Element ElmXnet is used to represent external networks. The
models used in the different calculation functions are described in the following chapters.

DIgSILENT PowerFactory 2023, Technical Reference 1

External Grid (ElmXnet)
2 Load Flow Model

2 Load Flow Model

The model used for the Load Flow calculation depends upon the selected bus type:

• PQ Bus Type
The external grid is modelled as a constant P (active power) and constant Q (reactive
power) infeed. A positive value for P is considered to be a generated active power and a
negative value is considered to be a consumed active power.
• PV Bus Type
The external grid infeeds a constant active power (for P > 0) and by default controls
the voltage of the busbar to which it is connected. Alternatively, the user may select a
reference busbar for voltage control. The voltage setpoint is defined in p.u. of the busbar
voltage. Only one external grid may control the voltage at any given busbar.
• SL Bus Type
The external grid controls the voltage, the angle and the frequency of the busbar to which it
is connected. If a reference busbar is selected, the voltage and the angle of this reference
busbar are controlled. In such cases, the user must define the voltage set point (in p.u.)
and the reference angle (in degrees).

Optionally, the active and reactive power input values may be entered as a combination of: P
and cos(phi), S and cos(phi), Q and cos(phi), S and P or S and P.

If the active power is set to zero and the option Out of service when active power is zero, the
machine does not take part in any voltage or frequency control.

2.1 Primary Frequency Control

A primary controlled state of a power system represents a state following an active power distur-
bance, in which the primary governors have settled and the system finds a “quasi steady-state”
before the secondary controlled power plants take over the active power balancing task. During
the primary frequency controlled state, the frequency deviates from nominal frequency. In or-
der to represent this state, the Load Flow calculation to be executed using active power control
according to primary control.

If the dispatched power of the participating elements is too low, in all isolated grids a common
frequency deviation dF in is calculated and the missing active power is distributed depending on
the primary frequency bias Kpf of the participating elements (power balance is established):

P = Pset − Kpf · dF in

where:

• P is actual active power in MW;

• Pset is active power setpoint in MW;
• Kpf is primary frequency bias in MW/Hz;
• dF in = F − F nom is frequency deviation in Hz.

Note: For SL Bus Type the frequency deviation is forced to 0. Hz (dF in = 0.).

An example of how the balancing active power is calculated and distributed to the participating
elements is shows in Figure 2.1.

DIgSILENT PowerFactory 2023, Technical Reference 2

External Grid (ElmXnet)
2 Load Flow Model

Figure 2.1: Example of a primary controlled Load Flow using Primary Frequency Bias factor

2.2 Secondary Frequency Control

For bringing frequency back to nominal frequency and/or for re-establishing area exchange flows
of an interconnected power system, secondary controlled power plants take over the active
power balancing task from the primary control after a few minutes (typically five minutes).

For simulating a secondary controlled state a power frequency controller needs to be specified
and the participating element needs to be part of this controller. The Load Flow calculation
needs to be executed using active power control according to secondary control. For more in-
formation related to the Power Frequency Controller object, refer to the corresponding Technical
Reference Manual.

If a Load Flow with active power control according to secondary control is calculated, the Ex-
ternal Grid participates to the secondary frequency control (as a part of a Power-Frequency
controller) and the active power is modified as:
P = Pset + (Kp · dP sco − K · dF in)
where:

• Pset is the active power setpoint in MW,

• K is the secondary frequency bias,
• dF in = F − F nom is the frequency deviation,
• Kp is the external grid participation factor for secondary frequency control, as defined in
the Power-Frequency controller,
• dP sco is the power unbalance (deviation), as calculated by the Power-Frequency con-
troller.

Normally the frequency deviation dF in is always zero for the secondary frequency control. This
is not the case when there is no frequency controller defined or if the participating elements
in the frequency controller are already in the limit. The corresponding calculation quantities
(signal) for the frequency deviation can be found in the variable selection dialogue (s : dF in in
Hz).

2.3 Operational Limits

The entered Reactive Power Operational Limits in the load flow page are considered in load
flow analysis only if the option Consider Reactive Power Limits is selected in the Load Flow

DIgSILENT PowerFactory 2023, Technical Reference 3

External Grid (ElmXnet)
2 Load Flow Model

calculation dialogue. Alternatively, PowerFactory allows defining the Capability Curve, which
can be manually entered. The P-Q characteristic of the capability curve is defined using either
p.u. values or MW/Mvar values. PowerFactory allows defining the Capability Curve considering
voltage dependent limits.

2.4 Unbalanced Load Flow

The calculation of the negative and zero sequence impedances used in the unbalanced Load
Flow calculation is carried out using the positive sequence reactance. The positive sequence
reactance is not used directly in the calculation for the load flow models. The maximum or the
minimum entered values are used for the calculation depending if the minimum or maximum
values option of the drop-down list Use for calculation is selected:
2
cf actor Unom
X1 = s · [Ω]
 2
R Sk00 (1)
1+
X

where :

• Unom is the nominal voltage of the busbar to which the external grid is connected,
• cf actor is the c-factor coefficient (minimum or maximum value depending on the selected
option Use for calculation),
• Sk00 is the short-circuit power (minimum value snssmin or maximum value snss),
R
• is the positive sequence R to X short circuit-ratio (minimum value rntxnmin or maxi-
X
mum value rntxn).

The zero sequence resistance and reactance are calculated as:

X0
X0 = · X1
X1 (2)
R0
R0 = · X0
X0
where :
X0
• is the short-circuit ratio between the zero and positive sequence reactance (minimum
X1
value x0tx1min or maximum value x0tx1),
R0
• is the zero sequence R to X short circuit-ratio (minimum value r0tx0min or maximum
X0
value r0tx0).

The negative sequence resistance and reactance are calculated as:

Z2 cf actor
Z2 = ·
Z1 Sk00
Z2
X2 = s  2 (3)
R
1+
X
p
R2 = Z22 − X22

where :

DIgSILENT PowerFactory 2023, Technical Reference 4

External Grid (ElmXnet)
2 Load Flow Model

Z2
• is the negative to positive sequence short-circuit impedance ratio (minimum value
Z1
z2tz1min or maximum value z2tz1).

DIgSILENT PowerFactory 2023, Technical Reference 5

External Grid (ElmXnet)
3 Short-Circuit Models

3 Short-Circuit Models

3.1 Short-Circuit Positive Sequence Model

The short-circuit positive sequence model is depicted in Figure 3.1.

Figure 3.1: Short-Circuit Positive Sequence Model

2 2
cf actor Unom cf actor Unom
X1 = q · 00 =q ·√ [Ω] (4)
R 2
1 + (X ) Sk 1 + (X R 2
) 3 · Ik00 · Unom

and
R
R1 = · X1 [Ω] (5)
X
where Unom is the nominal voltage of the busbar to which the external grid is connected.

3.2 VDE/IEC Short-Circuit Model

The method used to calculate the internal resistance and reactance is dependent upon the
00
short-circuit model used. For maximum short-circuits, the parameters Skmax and cf actor = cmax
00
are used. For minimum short-circuits, the parameters Skmin and cf actor = cmin are used. The
voltage factor, cf actor , depends on the voltage level of the busbar to which the external grid is
connected.

3.3 ANSI Short-Circuit Model

The ANSI short-circuit method defines Sk00 and cf actor using the corresponding input parameters
depending on whether the option “min. Values” or “max. Values” has been selected.

3.4 Complete Short-Circuit Model

The Complete short-circuit method defines Sk00 and cf actor using the corresponding input param-
eters depending on whether the option “min. Values” or ”max. Values” has been selected. The
pre-fault positive sequence model is depicted in Figure 3.2.

DIgSILENT PowerFactory 2023, Technical Reference 6

External Grid (ElmXnet)
3 Short-Circuit Models

Figure 3.2: Short-Circuit Pre-Fault Positive Sequence Model

The internal voltage depends on the load flow result and the pre-fault voltage and current, and
is calculated according to:

U0 = U1 pref + I1 pref · (R1 + jX1) (6)

U1 pref = U 1ldf (7)
I1 pref = I1ldf (8)

where U 1ldf is the positive sequence voltage and I1ldf is the positive sequence current after a
Load Flow calculation.

The entered Sk00 or Ik00 can be obtained from the element if the cf actor is set to 1 and in the Short-
Circuit Calculation dialogue the load flow initialisation is disabled and the all positive sequence
data is ignored.

3.5 Short-Circuit Zero Sequence Model

The short-circuit zero sequence model is depicted in Figure 3.3.

Figure 3.3: Short-Circuit Zero Sequence Model

DIgSILENT PowerFactory 2023, Technical Reference 7

External Grid (ElmXnet)
3 Short-Circuit Models

The zero sequence resistance and reactance are calculated according to following equations:
X0
X0 = · X1 (9)
X1
R0
R0 = · X0 (10)
X0

For ungrounded external grids, the grounding resistance (Re) and reactance (Xe) are neglected
(infinitive). Additionally, it is possible to connect a neutral conductor to the star point (see option:
External Star Point).

Assuming that Z1 = Z2 where ( Z2 R1 R0

Z1 = 1), X1 = X0 and the grounding impedance is zero
(Ze = 0), it is possible to calculate the X0/X1 ratio if the short-circuit current Ik00 (3p) for a three-
phase fault and the short-circuit current Ik00 (1p) for a single-phase to ground fault are given. This
ratio is calculated according to the following equation:
00
X0 3 · Ik3p
= 00 −2 (11)
X1 Ik1p

where the fault current for a three phase fault is:

cf actor · Unom
I 00k3p = (12)
Z1

and the fault current of a single phase to ground fault is:

3 · cf actor · Unom
I 00k1p = (13)
Z1 + Z2 + Z0 + 3 · Ze

3.6 Short-Circuit Negative Sequence Model

The short-circuit negative sequence model is depicted in Figure 3.4.

Figure 3.4: Short-Circuit Negative Sequence Model

The negative sequence resistance, R2, and reactance, X2, are calculated according to following
equations:
Z2
R2 = · R1 (14)
Z1
Z2
X2 = · X1 (15)
Z1

DIgSILENT PowerFactory 2023, Technical Reference 8

External Grid (ElmXnet)
4 Harmonics Model

4 Harmonics Model

4.1 Calculation of Network Impedance

The impedance of the external grid, which is taken into account for power quality assessment,
is calculated internally based on one of: the short-circuit power, Sk, at normal operation; the
maximum short-circuit power, Sk”max for faulted operation; or the minimum short-circuit power
Sk”min for faulted operation, depending on the user’s selection.

If the user has selected the network impedance to be based on the short-circuit power, Sk, and
impedance angle, psik, data for input fields SkV, psikV, z2tz1kV, x0tx1kV, and r0tx0kV can first
be calculated from a detailed network model using the Harmonic Load Flow command option
Calculate Sk at Fundamental Frequency, performed, for example, by the network operator. A
third party, (i.e. a wind farm planner) could get this information for the point of common coupling
(PCC for the planned wind farm) from the network operator. The planner can then enter the
data into the external grid element, which is a simplified representation of the network as seen
from the PCC.

4.1.1 Calculation of Positive Sequence Voltage

The positive sequence voltage, uini1 , is depicted in Figure 4.1.

Figure 4.1: Harmonics Positive Sequence Model

The positive sequence voltage is calculated as follows:

uini1 = u1 + i1 · z1 (16)

where z1 = R1 + jX1 and is calculated using values which are dependent upon the setting of
option ‘Used for calculation’ (cusedhrm):

1. If option ‘Max. Values’ has been selected:

p
X1 = cmax /snss/ 1 + rntxn2 (17)
R1 = rntxn · X1 (18)

2. If option ‘Min. Values’ has been selected:

q
X1 = cmin /snssmin / 1 + rntxn2min (19)
R1 = rntxnmin · X1 (20)

DIgSILENT PowerFactory 2023, Technical Reference 9

External Grid (ElmXnet)
4 Harmonics Model

3. If option ‘Sk’ has been selected:

1 π
X1 = · sin(psikV · ) (21)
SkV 180
1 π
R1 = · cos(psikV · ) (22)
SkV 180

Following the internal conversion of the calculated impedance values into units of Ohms, any
user-defined characteristics for R1 and L1 are applied to z1 (see the ‘Frequency Dependencies’
tab on the ‘Harmonics/Power Quality’ page of the external grid element).

4.1.2 Calculation of Negative Sequence Voltage

The negative sequence voltage, uini2 , is depicted in Figure 4.2.

Figure 4.2: Harmonics Negative Sequence Model

The negative sequence voltage is calculated as follows:

uini2 = u2 + i2 · z2 (23)

where z2 = R2 + jX2 and is calculated using values which are dependent upon the setting of
option ‘Used for calculation’ (cusedhrm):

1. If option ‘Max. Values’ has been selected:

X2 = z2tz1 · X1 (24)
R2 = z2tz1 · R1 (25)

2. If option ‘Min. Values’ has been selected:

X2 = z2tz1min · X1 (26)
R2 = z2tz1min · R1 (27)

3. If option ‘Sk’ has been selected:

X2 = z2tz1kV · X1 (28)
R2 = z2tz1kV · R1 (29)

where X1 and R1 are as defined in Section 4.1.1. Following the internal conversion of the
calculated impedance values into units of Ohms, any user-defined characteristics for R2 and
L2 are applied to z2 (see the ‘Frequency Dependencies’ tab on the ‘Harmonics/Power Quality’
page of the external grid element).

DIgSILENT PowerFactory 2023, Technical Reference 10

External Grid (ElmXnet)
4 Harmonics Model

4.1.3 Calculation of Zero Sequence Voltage

The zero sequence voltage, uini0 , is depicted in Figure 4.3.

Figure 4.3: Harmonics Zero Sequence Model

The zero sequence voltage is calculated as follows:

uini0 = u0 + i0 · (z0 + 3ze ) (30)

where z0 = R0 + jX0 and is calculated using values which are dependent upon the setting of
option ‘Used for calculation’ (cusedhrm):

1. If option ‘Max. Values’ has been selected:

X0 = x0tx1 · X1 (31)
R0 = r0tx0 · X0 (32)

2. If option ‘Min. Values’ has been selected:

X0 = x0tx1min · X1 (33)
R0 = r0tx0min · X0 (34)

3. If option ‘Sk’ has been selected:

X0 = x0tx1kV · X1 (35)
R0 = r0tx0kV · X0 (36)

where X1 is as defined in Section 4.1.1. Following the internal conversion of the calculated
impedance values into units of Ohms, any user-defined characteristics for R0 and L0 are applied
to z0 (see the ‘Frequency Dependencies’ tab on the ‘Harmonics/Power Quality’ page of the
external grid element).

4.1.4 Calculation of Zero Sequence Voltage Considering Neutral

The zero sequence voltage, uini0 , for an external grid with neutral, is depicted in Figure 4.4.

The zero sequence voltage, for an external grid with neutral, is calculated as follows:

uini0 = un − (i0 · z0 ) − u0 (37)

DIgSILENT PowerFactory 2023, Technical Reference 11

External Grid (ElmXnet)
4 Harmonics Model

Figure 4.4: Harmonics Zero Sequence Model Considering Neutral

where z0 is as defined in Section 4.1.3. Following the internal conversion of the calculated
impedance values into units of Ohms, any user-defined characteristics for R0 and L0 are applied
to z0 (see the ‘Frequency Dependencies’ tab on the ‘Harmonics/Power Quality’ page of the
external grid element).

4.2 Harmonic Voltages and Background Harmonics

Harmonic voltages may be entered according to one of the following two options:

• Phase Correct (where harmonic voltage magnitudes and angles may be entered for inte-
ger and non-integer harmonic orders)

• IEC 61000 (where harmonic voltage magnitudes may be entered for integer and non-
integer harmonic orders)

Background harmonics can be modelled using either the external grid element or the AC volt-
age source element, on their respective Harmonics/Power Quality pages. If only the harmonic
voltage amplitude is known (and not the angle), the IEC 61000 option can be selected.

The consideration of sequence component for the harmonic orders defined in the external grid
is detailed in Table 4.1.

DIgSILENT PowerFactory 2023, Technical Reference 12

External Grid (ElmXnet)
4 Harmonics Model

Table 4.1: Consideration of Sequence Components of Harmonic Injections for External Grid

Harmonic ElmXnet Setting Sequence Components of Harmonic Injections

Load Flow
Balanced
Phase Correct
• Positive (i.e. 4, 7, 10, ...), negative (i.e. 2, 5, 8, ...)
• Non-integer harmonic orders (i.e. 5.5, 6.2, 8.35,
...) are considered in the positive sequence
• Zero sequence orders are ignored (with warning)

IEC 61000
• Positive, negative

• Zero sequence orders (i.e. 3, 6, 9, ...) and non-

integer harmonics are in the positive sequence

Unbalanced
Phase Correct
• Positive, negative, zero
• Non-integer harmonics are considered

IEC 61000
• As for balanced harmonic load flow

4.3 Frequency Dependencies

Frequency dependent characteristics can be defined for the external grid. The definition of such
characteristics is described in the User’s Manual chapter Harmonic Analysis, section Frequency
Dependent Parameters, and Chapter Parameter Characteristics.

DIgSILENT PowerFactory 2023, Technical Reference 13

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

5 RMS- and EMT-Simulation

For the simulation model of the external network a model of a synchronous generator is being
used. The subtransient reactances of the model are predefined to 0.2p.u. and the saturation
and leakage rectances are neglected in the model so that the model takes a simplified form.

5.1 Parameters

The nominal power of the External Grid is calculated depending if maximum or minimum short-
circuit data is used:
 √ s00

 0.2 · 1 + rntxn2 · k M V A if maximum values are used

 cmax
snom = (38)
00
 p s
 0.2 · 1 + rntxn2 · kmin M V A if minimum values are used


min
cmin
where:

• s00k is the short circuit (snss) power in MVA;

• rntxn is the R to X ratio;

• c is the c-factor.

Similarly, the stator resistance is calculated depending if maximum or minimum short-circuit

data is used:

 0.2 · rntxn [p.u.] if maximum values are used
rs = (39)
0.2 · rntxnmin [p.u.] if minimum values are used


where the minimum value is limited to 1e−4 [p.u.]

The internal grounding impedance and neutral connection information can be defined in the
Grounding/Neutral Conductor tab of the Basic Data page of ElmXnet. The internal grounding
impedance per unit values are calculated as:
snom
rearth = Re ·
u2nom
snom (40)
xearth = Xe · 2
unom

where unom is the nominal voltage of the connected busbar.

The zero sequence reactance and resistance are calculated as follows:


 0.2 · x0tx1 [p.u.] if maximum values are used
x0 = (41)
0.2 · x0tx1min [p.u.] if minimum values are used



 x0 · r0tx0 [p.u.] if maximum values are used
r0 = (42)
x0 · r0tx0min [p.u.] if minimum values are used


DIgSILENT PowerFactory 2023, Technical Reference 14

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

For the unbalanced RMS-simulation, the negative sequence resistance is set equal to the sta-
tor resistance r2 = rs and using the pre-defined X and Y matrices, the negative sequence
reactance is calculated as:
 
1 Xd11 Xq 11
x2 = · + (43)
2 1 − Xd12 · Yd21 − Xd13 · Yd31 1 − Xq 12 · Yq 21 − Xq 13 · Yq 31

5.2 Flux equations and time constants

The rotor d-axis is modelled by two rotor loops representing the excitation (field) winding and
the damper winding. The q-axis is modelled using two damper windings. The model is written
using sub-transient variables.

The matrix of the d-axis and q-axis fluxes have the following forms:
       
ψd xd −1 1 id id
 ψe  = −(1 − σe ) · xd 1 −µe  ·  ie  = Xd ·  ie  (44)
ψD (1 − σD ) · xd −µD 1 iD iD

       
ψq xq −1 1 iq iq
 ψx  = −(1 − σx ) · xq 1 −µx  ·  ix  = Xq ·  ix  (45)
ψQ (1 − σQ ) · xq −µQ 1 iQ iQ

where µ is translation/screening factor and σ is decrement factor and saturation is neglected

and leakage reactance xl is zero.

These factors are calculated as follows:

x0d
σe =
xd
x0d · x00d
σD =
x0d · x00d + xd · (x0d − x00d )
xd − x0d
µe =
xd
xd · (x0d − x00d )
µD = 0
xd · x00d + xd · (x0d − x00d )
(46)
x0q
σx =
xq
x0q · x00q
σQ =
x0q · x00q + xq · (x0q − x00q )
xq − x0q
µx =
xq
xq · (x0q − x00q )
µQ = 0
xq · x00q + xq · (x0q − x00q )

DIgSILENT PowerFactory 2023, Technical Reference 15

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

The time constants of the rotor winding are calculated as:

t0d · xd
Te =
x0d
t00 · x0 x0 · x00 + xd · (x0d − x00d )
TD = d 00 d · d d
xd x002
d

(47)
t0q · xq
Tx =
x0q
t00q · x0q x0q · x00q + xq · (x0q − x00q )
TQ = ·
x00q x002
q

The stator and rotor currents can be expressed by using the inverse of the reactance matrices
Yd = Xd −1 and Yq = Xq −1 :
   
id ψd
 ie  = Yd ·  ψe  (48)
iD ψD

   
iq ψq
 ix  = Yq ·  ψx  (49)
iQ ψQ

5.3 Rotor equations

The rotor equations presented in this sub-section are used in both RMS and EMT simulations.

The following expressions are valid for the rotor currents:

ie = ψd · Yd21 + ψe · Yd22 + ψD · Yd23

iD = ψd · Yd31 + ψe · Yd32 + ψD · Yd33
(50)
ix = ψq · Yq 21 + ψx · Yq 22 + ψQ · Yq 23
iQ = ψq · Yq 31 + ψx · Yq 32 + ψQ · Yq 33

The rotor voltage equations for the d-axis and q-axis have the following form:

dψe ve − ie
=
dt Te
dψD iD
=−
dt TD
(51)
dψx ix
=−
dt Tx
dψQ iQ
=−
dt TQ

DIgSILENT PowerFactory 2023, Technical Reference 16

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

5.4 Stator equations

5.4.1 Stator equations used in the RMS-simulation

The stator fluxes in the d- and q-axis are calculated using the following equations:

ψd = x00d · id − ψd00
(52)
ψq = x00q · iq − ψq00

Using these definitions for the flux, the stator voltage equations can be written (with neglected
stator flux derivatives) as:
ud = u00d − rstr · id + n · x00q · iq
(53)
uq = u00q − rstr · iq − n · x00d · id

If the signal f ref is connected to the model then its value is used instead of the speed in
Equation 53.

The subtransient stator voltages in Equation 53 are calculated as:

u00d = −n · ψq00
(54)
u00q = n · ψd00

where the subtransient stator fluxes ψd00 and ψq00 are calculated as:

ψe · Yd12 + ψD · Yd13
ψd00 =
Yd11
(55)
ψ x · Y q 12 + ψQ · Yq 13
ψq00 =
Yq 11

The electrical torque te is calculated according to:

iq · ψd − id · ψq
te = [p.u.] (56)
cosn

In the case of unbalanced RMS simulation, additionally, the negative sequence, the zero se-
quence and neutral equations (if neutral is connected) have to be satisfied.

The negative sequence equations take into account the negative sequence impedance of the
model:

u2 = −(r2 +  · x2 ) · i2 (57)

Three different cases can be distinguished depending on the neutral conductor and internal
grounding impedance connection modes:

• No neutral connection and internal grounding impedance connected

In this case there is need only for zero sequence equations:

u0 = −(r0 +  · x0 ) · i0 − 3 · (rearth +  · xearth ) · i0 (58)

• N-connection at terminal (ABC-N)

When a neutral conductor is connected, zero sequence and equations for the neutral are
required. Here two sub-cases are possible:

DIgSILENT PowerFactory 2023, Technical Reference 17

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

– Internal grounding impedance not connected

u0 = −(r0 +  · x0 ) · i0 + un (59)
0 = 3 · i0 + in (60)

– Internal grounding impedance connected

u0 = −(r0 +  · x0 ) · i0 + un (61)
un = −(rearth +  · xearth ) · (3 · i0 + in ) (62)

• N-connection at separate terminal (internal grounding impedance is never connected)

u0 = −(r0sy +  · x0sy ) · i0 + un (63)

0 = 3 · i0 + in (64)

5.4.2 Stator equations used in the EMT-simulation

For the EMT model, the stator voltage equations are written using the stator αβ0 stationary
reference frame.

The following equations are obtained for the stator voltage uαβ = uα +  · uβ :

x00 diαβ x00 di∗αβ

uαβ = u00αβ − rstr · iαβ −  · 2 · n · x00∆ · e·2·ϕ · i∗αβ − · − ∆ · e·2·ϕ · (65)
ωn dt ωn dt
where the dq subtransient voltages are transformed to the αβ stationary reference frame using
the rotor position angle ϕ as u00αβ = (u00d +·u00q )·e·ϕ . The dq subtransient voltages are calculated
as:
Yq 13 1 dψD Yd13
u00d = u0d − n · ψQ · + · ·
Yq 11 ωn dt Yd11
(66)
Yd13 1 dψQ Yq 13
u00q = u0q + n · ψD · + · ·
Yd11 ωn dt Yq 11

where the transient voltages have the following form:

Yq 12 1 dψe Yd12
u0d = −n · ψx · + · ·
Yq 11 ωn dt Yd11
(67)
Yd12 1 dψx Yq 12
u0q = n · ψe · + · ·
Yd11 ωn dt Yq 11

x00d − x00q x00 + x00q

In the above equations, x00∆ and x00 are calculated as: x00∆ = and x00 = d .
2 2
The electrical torque te is calculated the same as in the RMS simulation (according to Equation
56).

Three different cases can be distinguished depending on the neutral conductor and internal
grounding impedance connection modes:

• No neutral connection and internal grounding impedance connected

In this case there is a need only for a zero sequence equation:
1 di0
u0 = −(r0 + 3 · rearth ) · i0 − (x0 + 3 · xearth ) · · (68)
hpi dt

DIgSILENT PowerFactory 2023, Technical Reference 18

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

• N-connection at terminal (ABC-N)

When a neutral conductor is connected, a zero sequence and an equation for the neutral
are required. Here two sub-cases are possible:
– Internal grounding impedance not connected

1 di0
u0 = −r0 · i0 − x0 · · + un (69)
hpi dt
0 = 3 · i0 + in (70)

– Internal grounding impedance connected

1 di0
u0 = −r0 · i0 − x0 · · + un (71)
hpi dt
 
xe di0 din
un = −re · (3 · i0 + in ) − · 3· + (72)
hpi dt dt

• N-connection at separate terminal (internal grounding impedance is never connected)

1 di0
u0 = −r0 · i0 − x0 · · + un (73)
hpi dt
0 = 3 · i0 + in (74)

5.5 Mechanical Equations

The speed derivative dn/dt of the machine is calculated using the following equation:

dn tm − te
= (75)
dt tag

where:

• tm is the mechanical torque in p.u.;

• te is the electrical torque in p.u. calculated according to Equation 56;

• tag is the acceleration time constant in s (also referred as the mechanical starting time and
it is equal to tag = 2 · H where H is the inertia constant).

From Equation 75 can be seen that, if there is a difference between the torques, the rotor will
be accelerated or decelerated. This equation represents the equation of motion.

The base torque is the ratio between the electrical active power and the nominal mechanical
angular frequency in mech. rad/s:
sgn · cosn
tbase = [p.u.] (76)
ω0m

The calculation of the mechanical torque tm (parameter xmt) is calculated as:

pt dp
tm = + + xmdm + addmt (77)
n n
where:

• pt is the Turbine Power input signal in p.u..

DIgSILENT PowerFactory 2023, Technical Reference 19

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

• dp is the output of internal speed controller;

• xmdm is the Torque Input input signal in p.u.;

• addmt is the Additional Torque parameter in p.u.. It can be used as an additional MDM
torque for the motor or as an additional pt torque for the generator.

The parameter dp is calculated using the droop which in turn is calculated using the secondary
frequency bias K:

dp = droop · (1 − n)
fnom (78)
droop = K ·
snom · cosn

5.5.1 Rotor angle definition

The angle ϕ (parameter phi) is the position of the rotor (d-axis) referenced to the reference
voltage of the network. The angle ϕ is a state variable in the model and its time derivative dϕ/dt
is calculated as follows:

ω · (n − fref ) [rad] in the RMS simulation
dϕ  n
= (79)
dt
ωn · n [rad] in the EMT simulation


where fref is an input signal connected to the reference machine frequency automatically by
PowerFactory.

In the RMS simulation, for the reference machine (slack), the reference frequency is equal to
the speed (fref = n), so the derivative of the angle is set to zero ( dϕ dt = 0). The state variable ϕ
is initialised by first calculating the q-axis angle and then shifting this angle for 90◦ :
π
ϕ = arctan (ut + (rs +  · xq ) · it ) − [rad] (80)
2
where ut is the terminal voltage of the machine, rstr is the stator resistance, xq is the q-axis
synchronous reactance of the machine and it is the current flowing through the machine.

5.5.2 Initialisation of the mechanical parameters

The mechanical torque is initialised with the value of the electrical torque. Since the initial speed
is 1p.u., pt is initialised with the value of the electrical torque. xmdm and addmt are initialised
to zero.

5.5.3 Excitation system interfacing

For interfacing with the excitation system, the excitation current output signal ie and the exci-
tation voltage input signal ve are used which are based on a non-reciprocal p.u. system. This
p.u. system can be also referred to as a no load, no saturation p.u.-system. Under no-load,
steady-state and rated speed conditions, excitation current of ie = 1 p.u. is required to produce
1 p.u. stator voltage on the air-gap line. Excitation voltage of ve = 1 p.u. is the corresponding
excitation voltage. The excitation voltage input signal is being initialised as:

ve = ie (81)

DIgSILENT PowerFactory 2023, Technical Reference 20

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

5.6 Using default machine input parameters

The default values of the synchronous, transient and sub-transient reactances, the transient and
sub-transient time constants in the d- and q-axis and the power factor are presented in Table
5.1.

Table 5.1: Default machine parameters in the d- and q-axis

Parameter Symbol Value Description

xd xd 0.2p.u. Synchronous reactance d-axis
xq xq 0.2p.u. Synchronous reactance q-axis
xds x0d 0.2p.u. Transient reactance d-axis
xqs x0q 0.2p.u. Transient reactance q-axis
xdss x00d 0.2p.u. Subtransient reactance d-axis
xqss x00q 0.2p.u. Subtransient reactance q-axis
tds t0d 1s Short-circuit transient time constant d-axis
tqs t0q 1s Short-circuit transient time constant q-axis
tdss t00d 0.1s Short-circuit subtransient time constant d-axis
tqss t00q 0.1s Short-circuit subtransient time constant q-axis

Using these default values, the model matrices get a simplified form, all the σ parameters obtain
the value 1 p.u. and all µ parameters obtain the value 0 p.u.. Similarly, the time constants in the
d- and q-axis obtain the values Te = Tx = 1 s and TD = TQ = 0.1 s. The negative sequence
reactance is obtained as x2 = 0.2 p.u.. Using the default input parameters, the dynamic model
of the external network is highly simplified.

5.7 Inputs/Outputs/State Variables of the Dynamic Model

5.7.1 Stability Model (RMS)

Table 5.2: Input Definition of the RMS-Model

Input Signal Symbol Description Unit

ve ve Excitation voltage p.u.
pt Turbine power p.u.
xmdm Torque input p.u.
fref fref Reference frequency p.u.

Table 5.3: State Variables Definition of the RMS-Model

Parameter Symbol Description Unit

psisr Stator flux, real part p.u.
psisi Stator flux, imaginary part p.u.
psie ψe Excitation flux p.u.
psiD ψD Flux in D-damper winding, d-axis p.u.
psix ψx Flux in x-damper winding, q-axis p.u.
psiQ ψQ Flux in Q-damper winding, q-axis p.u.
speed n Speed p.u.
phi ϕ Rotor position angle rad

DIgSILENT PowerFactory 2023, Technical Reference 21

External Grid (ElmXnet)
5 RMS- and EMT-Simulation

Table 5.4: Output Definition of the RMS-Model

Parameter Symbol Description Unit

uout Terminal Voltage, magnitude p.u.
xspeed Speed (xspeed = speed) p.u.

5.7.2 EMT-Model

Table 5.5: Input Definition of the EMT-Model

Input Signal Symbol Description Unit

ve ve Excitation Voltage p.u.
pt Turbine Power p.u.
xmdm Torque Input p.u.

Table 5.6: State Variables Definition of the EMT-Model

Parameter Symbol Description Unit

psie ψe Excitation flux p.u.
psiD ψD Flux in D-damper winding, d-axis p.u.
psix ψx Flux in x-damper winding, q-axis p.u.
psiQ ψQ Flux in Q-damper winding, q-axis p.u.
speed n Speed p.u.
phi ϕ Rotor position angle rad

Table 5.7: Output Definition of the EMT-Model

Parameter Symbol Description Unit

usd u0d Transient Voltage, d-axis p.u.
usq u0q Transient Voltage, q-axis p.u.
uout Terminal Voltage, magnitude p.u.
xspeed Speed (xspeed = speed) p.u.

DIgSILENT PowerFactory 2023, Technical Reference 22

External Grid (ElmXnet)
List of Figures

List of Figures

2.1 Example of a primary controlled Load Flow using Primary Frequency Bias factor 3

3.1 Short-Circuit Positive Sequence Model . . . . . . . . . . . . . . . . . . . . . . . . 6

3.2 Short-Circuit Pre-Fault Positive Sequence Model . . . . . . . . . . . . . . . . . . 7

3.3 Short-Circuit Zero Sequence Model . . . . . . . . . . . . . . . . . . . . . . . . . 7

3.4 Short-Circuit Negative Sequence Model . . . . . . . . . . . . . . . . . . . . . . . 8

4.1 Harmonics Positive Sequence Model . . . . . . . . . . . . . . . . . . . . . . . . . 9

4.2 Harmonics Negative Sequence Model . . . . . . . . . . . . . . . . . . . . . . . . 10

4.3 Harmonics Zero Sequence Model . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4.4 Harmonics Zero Sequence Model Considering Neutral . . . . . . . . . . . . . . . 12

DIgSILENT PowerFactory 2023, Technical Reference 23

External Grid (ElmXnet)
List of Tables

List of Tables

4.1 Consideration of Sequence Components of Harmonic Injections for External Grid 13

5.1 Default machine parameters in the d- and q-axis . . . . . . . . . . . . . . . . . . 21

5.2 Input Definition of the RMS-Model . . . . . . . . . . . . . . . . . . . . . . . . . . 21

5.3 State Variables Definition of the RMS-Model . . . . . . . . . . . . . . . . . . . . . 21

5.4 Output Definition of the RMS-Model . . . . . . . . . . . . . . . . . . . . . . . . . 22

5.5 Input Definition of the EMT-Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

5.6 State Variables Definition of the EMT-Model . . . . . . . . . . . . . . . . . . . . . 22

5.7 Output Definition of the EMT-Model . . . . . . . . . . . . . . . . . . . . . . . . . . 22

DIgSILENT PowerFactory 2023, Technical Reference 24

External Grid (ElmXnet)