See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/299424811

Static and Dynamic Calculation of Short-Circuit Currents in Synchronous

Generators

Conference Paper · June 2011

CITATIONS READS

3 6,636

5 authors, including:

Theofilos A Papadopoulos Panagiotis N Papadopoulos

Democritus University of Thrace University of Strathclyde
195 PUBLICATIONS 1,998 CITATIONS 70 PUBLICATIONS 594 CITATIONS

SEE PROFILE SEE PROFILE

Antonios Marinopoulos Grigoris Papagiannis

European Commission Aristotle University of Thessaloniki
56 PUBLICATIONS 871 CITATIONS 219 PUBLICATIONS 2,790 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Grigoris Papagiannis on 25 March 2016.

The user has requested enhancement of the downloaded file.

 Static and Dynamic Calculation of Short-Circuit
Currents in Synchronous Generators
T. A. Papadopoulos, Ch. G. Kaloudas, P. N. Papadopoulos, A. G. Marinopoulos, G. K. Papagiannis

Abstract— The calculation of short-circuit currents is mostly subtransient performance on SC currents is examined for
based on the methodology of the international standards different SC conditions and network topologies of isolated
IEC-60909 and ANSI / IEEE C37.010.1979. This methodology, and grid-connected generators. The effect of the fault location
although it utilizes simple procedures and various assumptions,
is also investigated and the analysis is extended to several
usually provides satisfactory results. However, dynamic tools
such as the ATP/EMTP can be used for the accurate simulation topologies with synchronous generators. Results are evaluated
of short-circuit currents. In this paper the influence of some using also the corresponding formulas for SC current
simplifying assumptions on the synchronous generator modeling calculations in synchronous generators [5].
in both standards is investigated and differences between the
static and the dynamic analysis are examined, highlighting cases II. SYSTEM UNDER STUDY
where significant discrepancies may occur. The analysis includes
single or multiple generators connected to isolated or grid The examined configuration is presented in Fig. 1 and
connected topologies. consists of two synchronous generators and two step-up
transformers connected through an Overhead Transmission
Keywords: Short-circuit calculation, IEC 60909, ANSI, Line (OHTL) to a busbar. The OHTL is of variable length A
ATP/EMTP modeling. and its remote end (N3) is connected either to a local load,
corresponding to an isolated network topology (Topology I),
I. INTRODUCTION or to a stiff busbar, representing generators operating at a

T HE calculation of the short circuit currents according to

the IEC-60909 [1] and the ANSI [2] standards is one of
the most common procedures for the power utility engineer.
grid-connected topology (Topology II). In both cases
generators are assumed to deliver rated power at nominal
power factor for the steady-state condition preceding the fault.
Both standards use a static calculation methodology, based on Generators and the wye side of the step-up transformers are
the equivalent voltage source method. This methodology can directly grounded.
be used for the calculation of maximum and minimum SC
currents, for both symmetrical and asymmetrical faults.
However, more precise results can be derived by the
calculation of SC currents from the actual waveforms that are
obtained by the dynamic simulation of the fault conditions.
The ATP/EMTP [3] software is a benchmark considering
transient studies in power systems and is certainly capable of
treating SC simulations [4].
In this paper, a thorough investigation of different SC
scenarios in networks with synchronous generators is
conducted. The methods of IEC-60909 and ANSI standards,
as well as ATP/EMTP dynamic simulations have been used, Fig. 1: System under study.
since the combined calculations and the comparison of the
results allow a better insight on the significance of all The above two network topologies are used to represent
assumptions, used in the SC calculations and useful two different grid schemes characterized as GS and GL. In
conclusions for practical engineering applications. scheme GS, two small hydro generators supply a 20 kV
More specifically, the influence of the generator medium voltage distribution busbar, while in scheme GL, two
large gas turbine generators are connected to a 150 kV
transmission busbar. The corresponding data are presented in
T. A. Papadopoulos, Ch. G. Kaloudas, P. N. Papadopoulos and G. K.
Papagiannis are with the Power Systems Laboratory, Department of Electrical
Table I.
and Computer Engineering, Aristotle University of Thessaloniki, Thessaloniki
GR 54124, Greece (e-mail: [email protected]). III. SYSTEM MODELING
A. G. Marinopoulos is with ABB Corporate Research, Västerås, Sweden.
The Short-Circuit calculation process according to the
Paper submitted to the International Conference on Power Systems IEC-60909 and ANSI standards is based on the equivalent
Transients (IPST2011) in Delft, the Netherlands June 14-17, 2011 voltage source method with certain assumptions. A thorough
analysis on the two static approaches, considering the prefault one the maximum peak current (ip-max) occurs, corresponding
conditions, the network topology representation and the SC to the maximum dc offset when the voltage is zero, while at
current calculations have been presented in [4], [7] – [9]. On this case the initial symmetrical current is symbolized as
the other hand, ATP/EMTP offers an adequate simulation of I''k@ip-max. In the second time the maximum initial
the dynamic behavior of the whole system during the fault. symmetrical current is calculated (I''k-max) and the
For the dynamic simulation, generators, transformers and the corresponding peak current is ip@I''k-max. This is illustrated in
OHTL are modeled using the Type 59 Synchronous Machine, Fig. 2a, where the three-phase SC current at N1 is presented
the BCTRAN and the JMarti models, respectively [3], while for different fault initialization times (tsc) and for the GS
the load is represented by an equivalent lumped R-L circuit. generator in network Topology II, assuming only one
generator in operation and the line length equal to 2 km. At tsc
TABLE I equal to 16 ms ip-max occurs, while at tsc equal to 18 ms I''k-max
TEST CASE NETWORK DATA
is recorded.
Data
Description Symbol
GL GS
(a) Different SC initialization times
Synchronous Generator 30

Rated Apparent Power (MVA) SG-r 161.7 1.35 t =16 ms

SC Current (kA)
20 sc
Rated Voltage (kV) VG-r 15 0.6 t =17 ms
sc
0.85 0.8 10 t =18 ms
Power Factor
sc
pf
lagging lagging
0
Synchronous Direct Axis
Xd 1.97 1.876
Reactance (pu) -10
0.01 0.02 0.03 0.04 0.05
Transient Direct Axis Reactance time (s)
X'd 0.22 0.211 (b) Initialization time at 16 ms
(pu) 30
ip-1
Subtransient Direct Axis SC Current (kA)
X''d 0.165 0.099 20
Reactance (pu)
SC-Transient Time Constant (ms) T'd 940 346 10

SC- Subtransient Time Constant

T''d 31 6 0
ip-2
(ms)
-10
Step-Up Transformer 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
time (s)
Rated Apparent Power (MVA) ST-r 180 1.6
Rated Voltage (kV) VT1/VT2 150/15 20/0.6 Fig. 2: Phase a current response during a three-phase SC at busbar N1 for the
Impedance Voltage (pu) uk 0.152 0.061 interconnected small generator scheme (a) different SC initialization times (b)
Resistive Comp. of Impedance current envelopes for initialization time at 16 ms.
ur 0.002 0.0113
Voltage (pu)
Vector Group - Yd5 Yd5 The initial symmetrical current in all cases is determined by
Overhead Transmission Line the top and the bottom envelopes of the SC current [11].
Positive Sequence OTL Therefore, I''k is computed using (1), as the sum or the
R'1+jX'1 0.09+j0.42 0.3+j0.3
Impedance (Ω/km)
difference of the two peak current values ip-1, ip-2 during the
Zero Sequence OTL Impedance 0.23+j1.31
(Ω/km)
R'0+jX'0 0.5+j1.7 first SC current cycle.
i p −1 ± i p − 2
Although in SC analysis accurate frequency domain I ''k = , (1)
2 2
modeling of the network elements is not necessary, the JMarti
model of ATP/EMTP has been selected in the analysis for the The sum in the numerator corresponds to the case where
OHTL. This has been done in order to avoid possible conflicts there is a zero crossing between ip-1 and ip-2, while the minus is
from the combination of the lumped PI equivalent model used when both peak currents have the same sign.
capacitances and the lumped load impedance. However in
both IEC 60909 and in the ANSI Standards simplified IV. INFLUENCE OF GENERATORS SUBTRANSIENT
transmission line models are used, ignoring the shunt PERFORMANCE
capacitances. Therefore in order to avoid errors due to the Both IEC-60909 and ANSI standards do not take into
different line models, another simulation has been conducted account the T''d and T'd time constants in SC calculations,
using a R-L equivalent for the transmission line. SC current which determine the subtransient and transient behavior of
results are compared to the corresponding by the JMarti line synchronous generators under fault conditions. In (A1) and
model showing very good agreement. (A3) of the Appendix the corresponding SC current formulas
Furthermore, in the dynamic simulation it is important to for the three- and single- phase to ground faults cases are
select the proper voltage conditions at the fault location, in presented, respectively, where the influence of T''d and T'd is
order to calculate the highest SC current. The maximum peak shown analytically.
and rms currents during a SC may occur at different times The influence of T''d on SC currents is investigated,
[10], thus two SC initialization times are examined. In the first examining three- and single- phase to ground faults at busbar
N1, using the ATP/EMTP software and the results are shown that the differences in the results for the Topology II
compared with the corresponding obtained by the IEC-60909 are 4 % to 6.5 % lower than the corresponding of Topology I.
and ANSI standards. Both network schemes GS and GL as well This is due to the different initial steady state conditions of the
as the two network topologies, for islanded network generator which are ignored in both standards [4], [7] – [9],
(Topology I) and for grid connected network (Topology II) while they are properly taken into account in ATP/EMTP.
are examined, assuming only generator G1 in operation and For the GS scheme per-cent differences may surpass 40 %,
the line length equal to 2 km. e.g. for I''k@ip-max current in Topology II. In the dynamic
simulation I''k is calculated using ip-1 and ip-2 of the top and
A. Three-phase to ground SC
bottom envelopes, as shown in Fig. 2b [1]. Current ip-1 is
SC currents calculated by the dynamic approach are shown 26.19 kA, while ip-2 is 0.34 kA, thus I''k is 9.14 kA. The initial
in Table II. The percent differences between SC currents (ISC), symmetrical current values calculated with the IEC and the
obtained by the two static approaches and by the dynamic ANSI Std. are 13.744 kA and 12.973 kA, respectively, and the
simulation, are calculated using (2). Differences for corresponding differences are -41.91 % and -50.37 %, given
Topology I and Topology II networks are presented in in Table IV.
Tables III and IV, respectively. Comparing the two generator schemes it is observed that
⎛ I SC − static procedure ⎞ for the GS scheme, in general, SC currents calculated by the
difference(%) = 100 ⋅ ⎜1 − ⎟⎟ , (2) dynamic simulation show higher deviation from the
⎜ I
⎝ SC − dynamic simulation ⎠ corresponding SC currents of the two standards, while these
differences for GL topology are significantly lower.
where, ISC can be ip-max, I''k@ip-max, ip@I''k-max, I''k-max. Observing This is attributed to the value of T''d, which is taken into
the results of Tables III and IV it must be noted that a minus account in dynamic simulation, allowing the exact recording
sign means that the fault current results by the corresponding of the subtransient performance of the generator. As shown in
method are higher than the reference. Table I, the value of T''d for the large generator is higher
compared to the corresponding of the small generator.
TABLE II
THREE-PHASE SC CURRENT AT BUSBAR N1 CALCULATED BY ATP/EMTP TABLE IV
Topology I Topology II % DIFFERENCES OF THREE-PHASE SC CURRENT AT NODE N1 FOR
SC Current SC time (ms)
(kA) (kA) TOPOLOGY II
Large Generator IEC IEC
15 108.0 113.84 ANSI
ip-max max. min.
I''k @ ip-max 15 38.18 40.25 Large Generator
ip @ I''k-max 15 108.0 113.84 ip-max -18.06 -22.82 -
I''k-max 15 38.18 40.25 I''k @ ip-max -23.68 -31.82 -19.84
Small Generator ip @ I''k-max -18.06 -22.82 -
ip-max 16 25.321 26.195 I''k-max -23.68 -31.82 -19.84
I''k @ ip-max 16 8.952 9.14 Small Generator
ip @ I''k-max 18 22.601 23.134 ip-max -33.72 -22.05 -
I''k-max 18 9.153 9.427 I''k @ ip-max -41.91 -50.37 -29.87
ip @ I''k-max -51.41 -38.20 -
TABLE III I''k-max -37.62 -45.79 -25.91
% DIFFERENCES OF THREE-PHASE SC CURRENT AT BUSBAR N1 FOR
TOPOLOGY I
Both I''k and ip are influenced by the subtransient
IEC IEC
ANSI
max. min. performance of the generator which takes place within the
Large Generator very first milliseconds of the fault. Therefore, higher T''d
ip-max -24.44 -29.47 - values result in longer subtransient times, which is a condition
I''k @ ip-max -30.36 -38.95 -26.32 more close to the methodology of calculation of I''k and ip SC
ip @ I''k-max -24.44 -29.47 - currents in IEC and ANSI standards. This is also illustrated in
I''k-max -30.36 -38.95 -26.32
Fig. 3, where exp(-t/Td''), defined in (A2b) assuming Xe equal
Small Generator
ip-max -38.33 -26.26 -9.04
to zero, is plotted for the GL and GS topologies, while Td''
I''k @ ip-max -44.91 -53.53 -32.59 takes different values. As Td'' increases, the term exp(-t/Td'')
ip @ I''k-max -54.98 -41.46 -22.17 acquires higher values, resulting in longer subtransient
I''k-max -41.74 -50.15 -29.68 periods, as occurs from (A1).

For the GL scheme ip-max and I''k-max occur for faults

occurring at 15 ms, while in GS scheme the corresponding
time is different, due to the lower value of T''d. Next,
comparing the results of the two network topologies, it is
 1
(a) Large Generator TABLE V
0.8
Td'' = 11 ms % DIFFERENCES OF SINGLE-PHASE TO GROUND SC CURRENT AT BUSBAR N1
Td'' = 31 ms
FOR THE SMALL GENERATOR TOPOLOGY
0.6 Td'' = 51 ms
IEC IEC ATP/
0.4 ANSI
max. min. EMTP
0.2
Large Generator
0
0 0.05 0.1 0.15 0.2 0.25 0.3 ip -10.27 -14.72 - 134.42 kA
time (s)
(b) Small Generator I''k -14.57 -22.11 -11.00 47.92 kA
1
Td'' = 3 ms Small Generator
0.8 Td'' = 6 ms
ip -19.18 -8.78 - 33.21 kA
0.6 T '' = 9 ms
d
I''k -1.47 -7.50 7.16 14.45 kA
0.4

0.2
In Table VI it is shown that the subtransient and transient
0
0 0.005 0.01 0.015 0.02 0.025
time (s)
0.03 0.035 0.04 0.045 0.05
time constants of the small generator for the single-phase to
Fig. 3: Values of exp(-t/Td'') term vs time for the (a) Large and (b) Small ground fault are higher than the corresponding for the
generators. three-phase to ground case, thus resulting in fault currents
which are closer to those calculated by the IEC and ANSI
30 approaches.
Td''= 3 ms
T ''= 6 ms
25 d TABLE VI
T ''= 9 ms
d TIME CONSTANTS OF THE SMALL GENERATOR FOR THE THREE- AND SINGLE-
20 PHASE TO GROUND SC
SC Current (kA)

Time Constant Three-Phase Single-Phase

15
Ta (ms) 21 19
10
Td'(ms) 346 564
Td''(ms) 6 9
5

V. INFLUENCE OF THE FAULT LOCATION

0
The effect of the electrical distance between the generator
-5
0 0.01 0.02 0.03 0.04 0.05 0.06 and the SC fault location is investigated. The GS Topology II
time (s)

Fig. 4: Phase a current responses during a three-phase SC at busbar N1 for the

scheme with only G1 generator in operation is considered,
interconnected small generator scheme and different T''d values. SC at 16 ms. while the fault location is assumed at busbars N1, N2 and N3
of the network in Fig. 1.
Current responses at busbar N1 for the different values of The per cent differences of SC currents for the IEC 60909
Td'' and network scheme GS operating in interconnected mode and ANSI standards are presented in Table VII for three- and
is presented in Fig. 4. The SC occurs at 16 ms, i.e. the case single- phase to ground faults. At the dynamic simulation the
where the maximum peak current is recorded. As T''d increases maximum values of SC currents ip and I''k occurred at
the first cycle peak current acquires higher values and the different times, thus the corresponding values are presented in
differences between the dynamic simulation results and the the grey outlined rows of the same table.
corresponding IEC SC currents are gradually reduced. The per TABLE VII
% DIFFERENCES OF THREE- AND SINGLE- PHASE TO GROUND SC CURRENTS
cent differences of I''k and ip compared to the corresponding FOR GS – TOPOLOGY II
maximum values of the IEC are -37.32 % and -14.39 %, Fault
N1 N2 N3
respectively. Position
The obtained results show that practically both IEC and Fault
3-ph 1-ph 3-ph 1-ph 3-ph 1-ph
ANSI standards treat synchronous generators with large T''d in Type
ip SC current
SC calculations more accurately than generators with small
ANSI
short circuit time constants. exact
-14.4 -0.21 -4.59 3.62 -3.15 5.91

B. Single-phase to ground SC ANSI

-33.72 -19.18 -23.93 -14.49 -23.73 -13.56
2.7 factor
Single-phase to ground SC current calculations are IEC -22.05 -8.77 -11.43 -2.98 -9.94 -1.07
presented in Table V for the scheme GS in Topology II. In this ATP
26.20 33.21 0.560 0.774 0.552 0.744
test case SC currents ip-max and I''k-max occur both for faults /EMTP (kA)
starting at 15 ms. I''k SC current
ANSI -37.62 -1.47 -15.77 3.81 -15.00 5.15
The corresponding differences of the SC currents acquire
IEC max. -45.79 -7.49 -22.52 -1.76 -21.81 -0.61
significantly lower values compared to the corresponding IEC min. -25.91 7.16 -11.71 7.33 -10.91 8.48
results of Table IV for the three-phase to ground fault. This is ATP
9.427 14.45 0.222 0.341 0.220 0.330
again attributed to the different value of T''d for the /EMTP (kA)
single-phase to ground fault, defined in (A4c).
 Considering the SC calculations with the ANSI Std., ip TABLE VIII
% DIFFERENCES OF THREE- AND SINGLE- PHASE TO GROUND SC CURRENTS
current is calculated using the coefficient derived by the actual
FOR TWO GL GENERATORS – TOPOLOGY II
X/R as well as by the approximate coefficient 2.7 [2]. The Fault
exact approach results in all cases to lower differences than N1/N4 N2 N3
Position
with the approximate approach. Differences between the Fault
3-ph 1-ph 3-ph 1-ph 3-ph 1-ph
ANSI Std. and the dynamic simulation for the single-phase to Type
ip SC current
ground case are practically negligible and have lower values
ANSI
than the minimum IEC SC currents in most cases, as shown in exact
9.46 1.55 12.92 3.80 13.56 4.93
Table VII. ANSI
11.44 5.73 14.92 4.35 14.65 6.15
2.7 factor
-3
7.6
x 10
IEC 4.36 -3.69 5.46 -5.21 6.09 -3.91
N3 (2km)
ATP
7.4 158.3 171.6 13.09 15.96 12.55 14.63
N3 (10 km)
/EMTP (kA)
7.2
I''k SC current
N2
ANSI 7.76 4.94 12.81 5.54 13.29 6.64
7 IEC max. -1.21 -4.00 2.87 -5.93 3.45 -4.49
IEC min. 7.99 5.45 11.71 3.70 12.24 5.02
T''de

6.8
ATP
56.28 64.38 4.73 5.89 4.57 5.45
6.6 /EMTP (kA)

6.4
N1 Fault locations of three– and single– phase to ground short
6.2 circuits are assumed at the busbars N1 – N4. SC currents
6
recorded at the faulted busbars N1 and N4 gave identical
0 0.01 0.02 0.03
Xe (pu)
0.04 0.05 0.06 0.07
results, due to the network electrical symmetry, thus the
Fig. 5: Three-phase equivalent subtransient time constant Tde'' against variable corresponding results are noted as N1/N4. In such complex
external impedance. network topologies it is difficult to analyze and calculate the
SC currents, using analytical formulas or an equivalent circuit
Absolute differences between the two static approaches and with a single equivalent time constant, as in the previous
the dynamic simulation are gradually decreased as the fault topologies [5].
location distance from the generator increases for both three- First, comparing the corresponding ANSI SC currents it is
and single- phase faults. This is due to the fact that the shown that for the GL generator scheme, the ANSI approach
equivalent subtransient time constant T''de is an increasing significantly underestimates the SC currents and especially ip
function with fault distance as shown in Fig. 5, where (A2b) is when calculated with the approximate coefficient 2.7. The
plotted against different values of the external impedance Xe, exact X/R ratio takes values from 43.5 to 62.5, depending on
thus resulting in longer subtransient performance periods. The the fault location and so is significantly higher than that
line length is assumed to be variable with length 2 and 10 km. assumed for the 2.7 coefficient approximation calculation [2].
For the three-phase to ground fault, SC current results by For the GS scheme, the ANSI approach significantly
the ATP/EMTP simulation are in all cases lower than the over-estimates the three-phase SC currents.
corresponding minimum SC currents of the IEC, while for the Results of the IEC ip and the maximum I''k SC currents for
single phase to ground fault, the dynamic simulation results the three-phase to ground fault for the GL scheme are slightly
are between the minimum and the maximum I''k and ip underestimated than the corresponding dynamic simulation
currents, according to IEC. Differences between the two static results, especially as the faulted node is electrically located
approaches and the dynamic simulation are more severe for further from the two generators. On the contrary for the GS
the three-phase to ground fault than for the single-phase to scheme, where the X/R ratio is significantly lower and takes
ground case. values from 5.3 to 6.1, the IEC method strongly overestimates
the corresponding three-phase to ground SC currents.
VI. MULTIPLE GENERATOR ARRANGEMENT For both network schemes I''k single-phase to ground SC
In this section the case of multiply fed short-circuits is currents calculated by the ATP/EMTP are between the
investigated, using the generalized network arrangement of corresponding maximum and minimum limits defined by the
Fig. 1, where both generators are connected to the network. IEC Std. In general the degree of the over- and
Schemes GL and GS are examined connected in Topology II underestimation of the two static approaches varies with initial
and the corresponding differences between the SC obtained by magnitude of each SC current source and the corresponding
the static approaches and the dynamic simulation are time constants [5] and therefore further systematic analysis is
presented in Tables VIII and IX, respectively. SC currents ip needed.
and I''k correspond to the maximum value recorded at the
corresponding time interval.
 generators. Overestimated and underestimated currents by the
TABLE IX two static approaches are recorded, depending on various
% DIFFERENCES OF THREE- AND SINGLE- PHASE TO GROUND SC CURRENTS
parameters and therefore general conclusion cannot be derived
FOR TWO GS GENERATORS – TOPOLOGY II
Fault and further systematic analysis is necessary.
N1/N4 N2 N3 The calculation procedures described in both IEC 60909
Position
Fault
3-ph 1-ph 3-ph 1-ph 3-ph 1-ph
and ANSI standards, are generally accepted as efficient
Type methods, leading to results that in most cases rely on the safe
ip SC current
side. However, there are certain cases, especially at the
ANSI
exact
-8.59 1.077 -4.50 3.61 -1.82 7.12 presence of multiple generators, where the user must be quite
ANSI careful to avoid overestimation or, even worse,
-31.81 -20.06 -23.82 -14.12 -23.71 -12.74
2.7 factor underestimation of the protection equipment. This paper
IEC -19.82 -2.76 -11.42 -1.61 -8.73 0.77 provides a better insight in the calculation of fault currents by
ATP
39.68 44.067 1.121 1.551 1.088 1.437 IEC, ANSI and dynamic simulation, highlighting cases where
/EMTP (kA)
such discrepancies may occur.
I''k SC current
ANSI -27.83 3.21 -16.03 4.09 -14.22 6.69
IEC VIII. APPENDIX
-35.85 -2.76 -23.02 -1.61 -21.33 0.78
max. In the generalized case where a three-phase SC occurs at a
IEC distant fault location form the generator terminals through an
-17.32 11.25 -11.74 7.60 -9.86 10.26
min.
external equivalent impedance Xe, the peak current envelope
ATP
15.15 20.25 0.443 0.684 0.436 0.643 at any time instant is given by (A1) [5].
/EMTP (kA)
⎡ 1 ⎛ 1 1 ⎞ − t Tde′ ⎤
VII. CONCLUSIONS iˆr ( t ) = 2 E0 ⎢ +⎜ − ⎟e ⎥
⎣⎢ X d + X e ⎝ X d′ + X e X d + X e ⎠ ⎦⎥
In this work different symmetrical three- and
⎛ 1 1 ⎞ − t Tde′′ 1
unsymmetrical single- phase to ground SC scenarios involving +⎜ − ⎟e + e − t Tae
synchronous generators operating either in an isolated ⎝ X d′′ + X e X d′ + X e ⎠ X d′′ + X e
, (A1)
network or in a grid-connected topology have been where:
investigated. Results obtained by the widely used ANSI and X ′ + Xe Xd
IEC 60909 standards and by dynamic simulation using the Tde′ = d × Td′
X d + X e X d′
ATP/EMTP software are compared in order to evaluate the , (A2a)
influence of various parameters. X d′′ + X e X d′
Tde′′ = × Td′′ , (A2b)
For the case of the grid-connected generator the short- X d′ + X e X d′′
circuit currents calculated by the two static approaches of the X d′′
Standards are closer to those obtained by the dynamic Ta = , (A2c)
ωs Ra
simulation, compared to the corresponding results for the
isolated network topology. Therefore, in isolated networks, as
in the islanded mode of operation, significant errors may and Ra is the stator dc resistance.
occur in the calculation of SC currents calculated by IEC- The corresponding equation for an unbalanced single-phase
60909 and ANSI/IEEE, in cases where generators are present. to ground SC at the poles of the generator is [5]:
It is shown that the influence of the generator subtransient ⎡ 1 ⎛ 1 1 ⎞ − t Td′(1φ ) ⎤
iˆr ( t ) = 3 2 E0 ⎢ +⎜ − NZ ⎟
e ⎥
time constant on the fault current magnitude is significant. ⎣⎢ X d + X e ⎝ X d′ + X e Xd + Xe ⎠
NZ NZ
⎦⎥
This parameter strongly affects the initial fault current
response. In both IEC 60909 and ANSI standards it is ⎛ 1 1 ⎞ −t Td′′(1φ ) 1 −t T
+⎜ − NZ ⎟
e + e a(1φ ) , (A3)
⎝ X d′′ + X e X d′ + X e ⎠ X d′′ + X e
NZ NZ
neglected, thus resulting in significant differences in the SC
current calculations especially for cases where synchronous where:
generators are characterized by very short subtransient time X NZ = X N + X Z , (A4a)
constants. Similarly the influence of the fault location is X′ + X NZ

examined. It is shown that as the electrical distance between Td′(1φ ) = d Tdo′ , (A4b)
X d + X NZ
the fault point and the generator increases, differences
between the two static approaches and the dynamic simulation X d′′ + X NZ
Td′′(1φ ) = Tdo′′ , (A4c)
are gradually reduced, since the equivalent subtransient time X d′ + X NZ
constant of the network increases, thus resulting in lower X d′′ + X NZ
subtransient periods. Ta (1φ ) = , (A4d)
ωs Ra
The investigation is extended for multiply fed faults,
examining the simultaneous operation of two synchronous and XN, XZ are the generator negative and zero sequence
impedances, respectively.

IX. REFERENCES
[1] Short-Circuit Currents in Three-Phase AC Systems, IEC Standard
60909. part 0 and 1, 1st edition, 2001-7.
[2] ANSI/IEEE Std 141, “IEEE Recommended practice for Electric Power
Distribution for Industrial plants”, Red Book, 1986.
[3] H.W. Dommel, EMTP Theory Book. Bonneville Power Administration,
Portland, OR, 1986.
[4] A. Berizzi, S. Massucco, A. Silvestri, “Short-Circuit Current
Calculation: A Comparison between Methods of IEC and ANSI
Standards Using Dynamic Simulation as Reference”, IEEE Trans. on
Industry Applications, Vol. 30, No 4., July/August 1994. Pp. 1099-1106.
[5] N. Tleis, Power Systems Modeling and Fault Analysis, Elsevier Ltd,
2008.
[6] C. N. Hartman, “Understanding asymmetry,” IEEE Trans. Industry
Applications, vol. IA-21, no. 4, pp. 267-273, July/Aug. 1985.
[7] G. Knight, H. Sieling, “Comparison of ANSI and IEC 909 Short-Circuit
Current Calculation Procedures”, IEEE Trans. on Industry
Applications, Vol. 29, no. 3, May/June 1993, pp. 625-630.
[8] Ch. G. Kaloudas, P. N. Papadopoulos, T. A. Papadopoulos, A. G.
Marinopoulos, G. K. Papagiannis, “Short-Circuit Analysis of an Isolated
Generator and Comparative Study of IEC, ANSI and Dynamic
Simulation,” presented at the MedPower 10 Conf., Agia Napa, Cyprus,
2010.
[9] A. J. Rodolakis, “A comparison of North American (ANSI) and
European (IEC) Fault Calculation Guidelines”, Industry Applications
Society Annual Meeting, Dearborn U.S.A., 1991.
[10] H. W. Reichenstein, J. C. Gomez, “Relationship of X/R, Ip and Irms' to
Asymmetry in Resistance/Reactance Circuits ,” IEEE Trans. Industry
Applications, vol. IA-21, no. 4, pp. 267-273, July/Aug. 1985.
[11] F. Castelli, A. Silvestri, D. Zaninelli, “The IEC 909 Standard and
Dynamic Simulation of Short-Circuit Currents”, ETEP, Vol. 4, No 3,
May/June 1994. Pp. 213-221.

View publication stats