SHORT-CIRCUIT CALCULATION WITH FULLSIZE CONVERTERS ACCORDING

TO IEC 60909

Gerd Balzer
Darmstadt University of Technology
Landgraf-Georg-Str. 4, D-64283 Darmstadt, Germany, [email protected]

ABSTRACT

Due to the increasing integration of wind farms and PV systems in the electrical networks, it is necessary to
consider the influence of these power plants on the short-circuit currents in transmission and distribution net-
works. In many cases, both the wind farm as well as the photovoltaic systems will be connected via VSC con-
verters (full-size converters) with the power system. As the old IEC Standard (IEC 60909-0: 2001-07: Short-
circuit currents in three-phase a.c. systems – Part 0: Calculation of currents) does not included the calculation of
currents generated by self-commutated converters, a revision of this standard has been necessary. In the new
version of IEC Standard Short-Circuit Calculation, the short-circuit current at the fault location is calculated in a
first step, which is caused by the voltage sources, and the VSC converters are modeled with an infinite internal
resistance in the positive sequence. In a second step, the existing voltage sources are shorted circuit, and the
current contribution of the converters are determined to the short-circuit current at the fault location by using
current sources. Subsequently, the superimposition of the two current components is carried out in order to de-
termine the total current. Moreover, it is possible to determine the voltage distribution throughout the network.
The maximum inductive current that the VSC converter can feed during the short duration must be provided by
the manufacturer. All necessary currents can be determined based on the initial short-circuit alternating current,
which are relevant for the design of equipment and installations: Peak short-circuit current, breaking short-
circuit current, and steady state short-circuit current.

KEYWORDS: Short-circuit calculation, full-size converter, initial short-circuit current, renewal energy

According to the new IEC 60909-0 [1], the short-circuit current calculation can also be conducted in case of
1 INTRODUCTION

full-size converters, e.g. if the connection of wind and PV systems are considered. Figure 1 shows an example
of the system circuit for these two cases. While the consideration of networks, generators and motors, uses volt-
age sources, a full-size converter has to be represented by a current source, which in some circumstances may
feed in the (inductive) short-circuit current (voltage controlled). In this report it is shown, how the current calcu-
lation can be performed at the fault location in this case according to the new IEC standard, exemplarily in case
of a three phase short circuit.

Figure 1: Installation of self-commutated converter

a: Wind power generation; b: Photo voltaic power generation

2 SIMULATION OF A POWER GENERATION WITH FULL-SIZE CONVERTERS

According to the new IEC standard, the full-size converter has to be simulated using the following circumstanc-
es:
o Power generation feeders with full-size converters (wind, photovoltaics) are modeled by a current source
in the positive sequence. The current source depends on the short-circuit failure, and the value will be
provided by the manufacturer. The shunt impedance of the power source ZPF is assumed to be infinite.
o In case of unbalanced short circuits, the shunt impedance of the negative sequence system Z(2)PF depends
on the fault type as well as the control strategy. The zero sequence impedance Z(0)PF can be considered as
infinite.
o Power generation feeders can be neglected, if their contribution to the total initial short-circuit current
without full-size converters is less than 5 %.

In the following sections, the equations for the three-phase short circuit are generally derived, in addition, the
three-phase short circuit is calculated for a 110 kV system.

3 CALCULATION OF THE THREE PHASE IITIAL SHORT-CIRCUIT CURRENT

3.1 Procedure
As an example, the system circuit is used according to Fig. 2. In total, three wind parks are connected to the 110
kV network, which is supplied by the upstream system.

Figure 2: Equivalent circuit for the calculation of the short-circuit current

The following data of the components are assumed for the calculation of an example,
Network
I k" = 20 kA
o

R/X = 0,1
Initial short-circuit current:
R/X ratio:

Z 'L = (0,120 + j0,393)Ω / km

o Overhead lines
impedance:
line lengths: 12 = 100 km , 13 =  23 = 50 km ,  34 = 25 km

PU 2 = 100 MW , PU 3 = 50 MW = PU4
o Full-size converter (wind park)
rated power:
short-circuit current: I kU = k ⋅ I rU , k = 1,2
According to Fig. 2, the 110 kV system is assumed to be a voltage source, while the full-converters represent a
constant current source. For this reason, in total two steps are used to calculate the short-circuit currents, namely:

o Calculation of short circuit current at the fault location I "kIi (busbar 2), caused by the 110 kV system
and
o calculation of short circuit current at the fault location I "kIIi (busbar 2), caused by the injection of the
full –size converters.
The two parts of the short-circuit currents I "kIi and I "kIIi are finally superimposed to determine the total short
"
circuit current I ki .

The equivalent circuit of the positive sequence of the three-phase short circuit is shown in Fig. 3, in this case
both the current sources of full-size converter and the equivalent voltage source at the fault location F2 are de-
picted.
 Figure 3: Equivalent circuit of the positive sequence system for the three phase short circuit at the location F2

3.2 Step 1: Equivalent voltage source at the fault location

Basically, the admittance matrix YN of the network is set up, followed by the calculation of the currents Ii at the
system node using the impedance matrix ZN and the voltage distribution Ui. In this connection, the current
sources are considered to be infinite:
Ii = Y N ⋅ Ui Ui = Z N ⋅ Ii (1)

Generally, the admittance resp. impedance matrix of a network becomes:

 Y 11  Y 1m   Z 11  Z 1m 
   
 Y 1i  Z 1i
       
Y N =  Y i1  Y im  Z N =  Z i1  Z im 
     

   
 Y ii  Z ii (2)
       
Y  Y mm  Z  Z mm 
     
 m1  Y mi  m1  Z mi

Fig. 4 applies to the equivalent voltage source at the fault location, with the amplitude of the equivalent voltage
source according to equation (3).

U i = c ⋅U n / 3 (3)

Figure 4: Equivalent circuit with the equivalent voltage source at the fault location

Using the impedance matrix, it is possible to calculate the short-circuit current at the fault location. Here, the
equivalent voltage is applied at node i according to equation (3), while all currents Ii of the system nodes are
equal to zero, except for the fault current. Generally, the equation becomes:
 U 1   Z 11 Z 1m   0 
     
 Z 1i 
         
U  = Z  Z im  ⋅  I "kIi 
  

 i   i1   
 Z ii (4)
         
U   Z  Z mm   0 
  
 m   m1  Z mi
 "
From equation (4), the short-circuit current I kIi results without the influence of full-size converters to:

c ⋅Un / 3
c ⋅ U n / 3 = Z ii ⋅ I "kIi I "kIi = (5)
Z ii

Subsequently, all node voltages related to ground can be derived from equations (4), so that, for example, the
voltage U1 at the system node 1 applies in case of a short circuit at node i:

U 1 = Z 1i ⋅ I "kIi = ⋅ c ⋅U n / 3
Z 1i
(6)
Z ii

It should be noted that the voltage according to equation (6) is obtained as a result of the changed voltage profile
due to the equivalent voltage source at the fault location. The actual voltage U1r becomes:

 Z 
U 1r = c ⋅ U n / 3 − ⋅ c ⋅ U n / 3 = c ⋅ U n / 3 ⋅ 1 − 1i 
Z 1i
 Z ii 
(7)
Z ii

3.3 Step 2: Constant current source

The representation of the full-size converters is performed by a current source with the internal impedance ZUi
→∞, so that the equivalent circuit is based on Figure 5.

Figure 5: Representation of self-commutated converter as current source (positive sequence system)

In a similar way, according to the method of the equivalent voltage source at the fault location (equation 4), the
current distribution is calculated exclusively by the influence of full-size converters, so that the system of equa-
tions is considered according to equation (8). At the fault location, the voltage Ui is short circuited, while the
remaining voltages at the system nodes are calculated due to the current conditions.

 U 1   Z 11 Z 1m   0   I kU1 
       
 Z 1i 
             
 0 = Z Z im  ⋅  I "kIIi  −  I kUi 
  

   i1     
 Z ii  (8)
             
U   Z Z mm   0   I kUm 
  
 m   m1  Z mi 

"
As a result of the power feed-in of the full-size converters, the short-circuit contribution I kIIi at the fault loca-
tion becomes:

0 = Z ii ⋅ I "kIIi − (Z i1 ⋅ I U1 +  + Z im ⋅ I kUm ) = Z ii ⋅ I "kIIi − ∑ Z ij ⋅ I kUj

m
(9)
j

After conversion:

I "kIIi = ⋅∑Z ⋅I
1 m
(10)
Z ii j ij kUj

3.4 Step 3: Superposition

The two current components according to the equations (5) and (10) are added together to calculate the total
current at the fault location, so that the following equation (11) applies.
 c ⋅U n / 3
I "ki = I "kIi + I "kIIi = + ⋅∑Z ⋅I
1 m
(11)
Z ii Z ii j ij kUj

The values Zij Zii are derived from the impedance matrix according to equation (2), where i represents the short-
circuit location and j the node, which the full-size converter is connected to the system. Complex values are

Standard [1].
used to determine the total current, whereas the amounts of the currents are considered according to the IEC

3.5 Step 4: Calculation of the voltage distribution

Starting from the current distribution, the voltages can be determined at the network nodes. Here, the system of
equations (12) is used taking the entire short-circuit current I "ki according section 3.3 into consideration.
 U 1   Z 11 Z 1m   0   I kU1 
       
 Z 1i 
             
U = Z  Z im  ⋅  I "ki  −  I kUi 
  

 i   i1     
 Z ii (12)
             
U   Z  Z mm   0   I kUm 
  
 m   m1  Z mi

Ui is the equivalent voltage source at the fault location according to equation (3). Even in this case, the calculat-
ed voltages represent the voltage changes due to the short circuit. The real voltages Uir at the system nodes are
calculated according equation (7).
3.6 Example
Based on the impedances of the equipment specified in section 3.1, the corresponding matrices are set up to
determine the initial short-circuit current according to equation 11. The following impedances can be calculated
due to the rated values (section 3.1):
110 kV system:
c ⋅ U nQ 1,1 ⋅ 110
ZQ = = ⋅ Ω = 3,4930 Ω Z Q = (0,3476 + j3,4756) Ω
3 ⋅ I k" 3 ⋅ 20

Z L12 = (12,000 + j39,300)Ω Z L13 = (6,000 + j19,650)Ω = Z L12 / 2

Overhead lines:

Z L 23 = (6,000 + j19,650) Ω = Z L12 / 2 Z L34 = (3,000 + j9,825) Ω = Z L12 / 4

Full-size converters (wind parks)

I rU2 = = kA = 0,5249 kA I rU3 = 0,2624 kA = I rU4

PU2 100
3 ⋅U n 3 ⋅ 110
As the injected current of the full-size converts is assumed to be inductive, the respective currents are arising in
case oof a short circuit to:
I kU2 = − j0,6298 kA I kU3 = − j0,3149 kA = I kU4

The admittance matrix can be derived according equation (2) for the example of Fig. 2.

 Y Q + Y L12 + Y L13 − Y L12 − Y L13 

 
0
 − Y L12 Y L12 + Y L 23 − Y L 23 0 
YN = 
− Y L13 − Y L 23 Y L13 + Y L 23 + Y L34 − Y L34 
 
(13)
 − Y L34 Y L34 
 0 0

Due to the considered values, the equation becomes:

 + 0,04981 − j0,35470 − 0,00711 + j0,02328 − 0,01421 + j0,04655 
 
0
 − 0,00711 + j0,02328 + 0,02132 − j0,06983 − 0,01421 + j0,04655 
YN = 
+ 0,05685 − j0,18620 − 0,02843 + j0,09310 
0
− 0,01421 + j0,04655 − 0,01421 + j0,04655
 
 0 0 − 0,02843 + j0,09310 + 0,02843 − j0,09310 
The inversion of the admittance matrix leads to the impedance matrix to calculate the short-circuit currents.
 + 0,3476 + j3.4756 + 0,3476 + j3.4756 + 0,3476 + j3.4756 + 0,3476 + j3.4756 
 
 + 0,3476 + j3.4756 + 6,3469 + j23,1238 + 3,3473 + j13,2997 + 3,3473 + j13,2997 
=
+ 0,3476 + j3.4756 + 3,3473 + j13,2997 + 4,8468 + j18,2132 + 4,8468 + j18,2132 
ZN
 
 + 0,3476 + j3.4756 + 3,3473 + j13,2997 + 4,8468 + j18,2132 + 7,8471 + j28,0381
 

The calculation on the basis of the example is carried out with complex quantities in accordance with equation
(11), in contrast to the IEC standard. The total short-circuit current at the location F2 (i = 2, busbar 2) becomes:

⋅ [Z 21 ⋅ I kU1 + Z 22 ⋅ I kU 2 + Z 23 ⋅ I kU 3 + Z 24 ⋅ I kU 4 ]
c ⋅U n / 3
I "kF2 = +
1
(14)
Z 22 Z 22

As no full-size converter feeds in at busbar (1) (IkU1 = 0) , the short-circuit current is calculated by simplifica-
tion:

⋅ [Z 23 ⋅ I kU 3 + Z 24 ⋅ I kU 4 ] + I kU 2
c ⋅U n / 3
I "kF2 = +
1
(15)
Z 22 Z 22

After insertion of the numerical values, equation (12) becomes with Z23 = Z24 and IkU3 = IkU4:

1,1 ⋅ 110 / 3 3,3473 + j13,2997

I "kF2 = kA − j ⋅ 0,6298 kA − j0,6298 kA = 0,7788 kA - j3,7994 kA
6,3469 + j23,1238 6,3469 + j23,1238

2 = 2,9133 kA ), while the

"
The first part relates to the contribution, which is caused by the 110 kV network ( I kIF
2 = 0,9899 kA ). The amount of the
"
second and third part is related to the feed-in of the full-size converter ( I kIIF
total current is:

2 = 3,8782 kA
"
I kF

If the current is calculated according to the IEC Standard with the amounts of the currents, the result becomes:

1,1 ⋅ 110 / 3
2 = kA + ⋅ 0,6298 kA + 0,6298 kA = 2,9134 kA + 0,9900 kA = 3,9034 kA
" 13,7144
I kF
23,9790 23,9790
In this case, the calculation with amounts provides a result which is greater by 0.65% as the value with complex
values. The voltage distribution at the network nodes may be determined with the aid of the equation system
(16).

 U 1   Z 11 Z 14   0   0 
       
Z 12 Z 13
U 2   Z 21 Z 24   I "kF2   I kU2 
U  =  Z ⋅ −
Z 34   0   I kU3 
Z 22 Z 23

 3   31     
U   Z Z 44   0   I kU4 
Z 32 Z 33
 4   41
(16)
Z 42 Z 43

The single voltage nodes can be considered taking into account the short-circuit current according to equation

[ ]
(15):

U1 = ⋅ c ⋅ U n / 3 + 2 ⋅ (Z 23 − Z 22 ) ⋅ I kU 3
Z 12
Z 22
U 2 = c ⋅U n / 3
Z 23   Z ⋅Z  
U3 = ⋅ c ⋅ U n / 3 + 2 ⋅  Z 23 − 22 33  ⋅ I kU 3 
(17)
Z 22   Z 23  
Z 23   Z ⋅ (Z 33 + Z 44 ) 
U4 = ⋅ c ⋅ U n / 3 + I kU3 ⋅  2 ⋅ Z 23 − 22 

Z 22   Z 23 
The voltages with full-size converters are calculated taking into account the values of the impedance matrix, in
this case the real distribution of the voltages Uir is considered according to equation (7):

U1r = 0,8702 ⋅ c ⋅ U n / 3 U 2 r = 0,0

U 3r = 0,5250 ⋅ c ⋅ U n / 3 U 4r = 0,5703 ⋅ c ⋅ U n / 3

If the full-size converters are not included, the node voltages become:

U1r = 0,8567 ⋅ c ⋅ U n / 3 U 2r = 0,0

U 3r = U 4 r = 0,4284 ⋅ c ⋅ U n / 3

The node voltages arise by inductive feed-in of the full-size converter during the short-circuit phase, this is true
of course not for the short-circuit location.
4 PEAK SHORT-CIRCUIT CURRENT
4.1 Calculation of the peak short-circuit current
In each case, the peak short-circuit current is superposed of two parts (without and with full-size converter) to
get the total current. The peak short-circuit current of a full-size converter is calculated according to equation
(18) to:

ip = 2 ⋅ I kU
"
(18)

With
"
I kU (= IsU) short-circuit current contribution of a full-size converter on the high voltage side, value will be
provided by the manufacturer, equation (10).
In a radial network, the resulting peak short-circuit current is calculated by the sum the partial currents. Assum-
ing a mesh network, in contrast to equation (18), the resulting current is obtained at any network node by equa-
tion (19):

ip = k ⋅ 2 ⋅ I kI
"
+ 2 ⋅ I kII
"
(17)

With
"
I kI three phase initial short-circuit current at the fault location without contribution of the full-size con-
verter according equation (5).
"
I kII contribution of the full-size converter to the three phase short-circuit current at the fault location,
equation (10).
When calculating the peak short-circuit current according to equation (19), a DC component of the current with-
out the full-size converter is taken into considerations (k factor), but not when determining the contribution of
the full-size converter. The example of Figure 6 shows the transient short-circuit current curve in case of a short
circuit in the 380 kV network. Here, a three-phase fault at the end of a short transmission line is assumed, which
is initiated at the zero crossing of voltage phase R.

Figure 6: Equivalent circuit for the calculation of the three phase short-circuit current of a HVDC installation

Figure 7 shows the current transients of conductor R of the network feeder (Fig. 7a) and the HVDC part (Fig.
7b). In addition, the RMS value is listed, which is calculated from the three phase voltages.
 a

Figure 7:Current transient of phase R in case of a three phase short circuit

a: System contribution (N); b: HVDC contribution (U)

After short circuit initiation, the current contribution of the HVDC network is controlled so that the rated current
is fed-in as an inductive current. Figure 8 shows the superposition of the current transients in each phase R. This
inductive current is superimposed to the network part at the fault location, which is predominantly inductive
current, due to the R/X ratio.
 Figure 8: Superposition of the transients phase R
iNR(t): Network; iUR(t):HVDC; iNUR(t): total current

The following results can be obtained from Figures 7 and 8:

o The fault current is primarily determined by the network part, which results from the ratio of the currents
I k' ' / I rU .
o The part of the full-size converter has no DC component, so that equation (19) is justified.
In this example, the maximum value of the fault current is increased from ip = 43,622 kA to ip = 44,324 kA,
which represents an increase of 1,609%.

4.2 Calculation of the peak short-circuit current (example)

The peak short-circuit current is calculated according to equation (19) considering the two parts of the initial

circuit consists of impedances with the same R/X ratio), and in consequence, the factor k can be obtained from
short-circuit current. In case of a short circuit at node 2 of Fig. 2, the system is a radial network (the parallel

the short-circuit impedance at the fault location. The resulting short-circuit impedance Z22 is derived from the
impedance matrix.

Z 22 = (6,3469 + j23.1238)Ω R / X = 0,2745 k = 1,4501

The peak short-circuit current becomes:

ip = 1,4501 ⋅ 2 ⋅ 2,9133 kA + 2 ⋅ 0,9899 kA = 5,9746 kA + 1,3999 kA = 7,3746 kA

5 BREAKING AND STEADY STATE CURRENT

The three phase breaking current as well as the steady state current are equal to the initial short-circuit current.
6 CONCLUSION
With the new IEC Standard, it is possible to calculate the short-circuit contribution of VSC converters (full-size
converters), which connect the wind and PV systems to the distribution and transmission grid. In this report,
only three-phase short-circuit current are presented, but the new IEC standard covers all types of short circuits.
The next step is to take the contribution of HVDC systems to the network short-circuit current into account
using a similar calculation pattern.
REFERENCES
[1] IEC 60909-0:2016 Edition 2.0 (2016-01-28): Short-circuit current calculations in three-phase a.c. systems –
Part 0: Calculation of currents