Analysis and Simulation of Causes of Voltage Sags

Using EMTP

Dung Vo Tien, Radomir Gono Zbigniew Leonowicz

VSB-TUO Technical University of Ostrava Wroclaw University of Science and Technology
Ostrava, Czech Republic Wroclaw, Poland
Dung.vo.tien.st@vsb.cz, radomir.gono@vsb.cz zbigniew.leonowicz@pwr.edu.pl

Abstract— This paper presents an inclusive set of EMTP load terminals) can be found from:
models used to simulate the cause of voltage sags such as short
circuits, transformer energizing, induction motor starting. ZF
Vsag = E (1)
Voltage sag is usually described as characteristics of both ZF + ZS
magnitude and duration, but it is also necessary to detect phase
angle jump in order to identify sags phenomena and finding the
Where ZS is the source impedance at the PCC and ZF is the
solutions, especially for sags due to short circuits. In case of the impedance between the PCC and the fault [6].
simulation of voltage sags due to short circuit, their effect on the
magnitude, duration and phase-jump are studied.

Keywords—voltage sag, power system, simulation model,

EMTP/ATP software.

I. INTRODUCTION
Voltage sags is one of the most common events that affect
power quality in the distribution system. Although voltage sags Fig. 1. Voltage divider model for a voltage sag.
have a short time durtion (from 0.5 cycles to 1 minute) [1], it
may impact on the operation of equipment (stalling of motors, Magnitude of the voltage sag can be determined in the
tripping of sensitive loads and inaccuracy of control devices). RMS voltages. According to the definition of root mean value,
Mitigating the voltage sag is necessary to determine the cause the RMS voltage over one data window can be calculated from
of the sag. With the development of computer technology and the time-domain voltages, this is done by using the following
software, analysis and simulation the causes of voltage sags equation:
can be made. There are many power system simulation tools
available, such as PSCAD/EMTDC [2], EMTP/ATP [3], 1 N 2
MATLAB with Power System Toolbox [4], [10], [11]. In this Vrms = ¦ vi (2)
N i =1
paper, ATP/EMTP is chosen as the simulation platform
because according to [5], ATP-EMTP is designed to simulate where N is number of samples per cycle and vi are the
the physical processes of transmission lines and transformers sampled voltages in time domain.
quickly and in a convenient way. For any other window length the RMS voltage calculation
Magnitude and duration are the main characteristics of the is typically made once a cycle:
voltage sag, the magnitude is defined as the reduction of
voltage during the sag and the duration is defined as the time 1 i =kN
Vrms (kN ) = ¦ vi2 (3)
between the sag starting and clearing. However, the controller N i =( k −1) N +1
using power electronic converter need phase angle information Where k is integer.
for compensating the voltage may be affected by the phase
angle jumps. Therefore, it is necessary to identify A shorter window than one half-cycle is not useful.
characteristics of magnitude, duration and phase angle
variations to finding the efficient solutions. B. Phase angle jump
From (1), let Z S = RS + jX S and Z F = RF + jX F and
II. VOLTAGE SAGS CHARACTERISTICS
assume E= 1, the argument of Vsag, thus the phase jump in the
A. Voltage sag magnitude voltage is given by the following expression.
Fig. 1 shows the simple voltage divider model. The voltage XF X + XF
at the point of common coupling (PCC) (also the voltage at the Δφ = arctan( ) − arctan( S ) (4)
RF RS + RF
This research was partially supported by the SGS grant from VSB -
Technical University of Ostrava (No. SP2017/54) and by the project
TUCENET (No. LO1404).

978-1-5386-3917-7/17/$31.00 ©2017 IEEE

 XS XF B. Single line to ground fault
If = expression (4) is zero and there is no phase
RS RF Single line to ground faults is the most popular fault in
angle jump. The phase-angle jump will thus be present if the power system, (about 80% of the total faults in power systems
X/R ratios of the source and the feeder are different [6]. of Vietnam). The simulated voltage sags waveform caused by a
single line to ground (SLG) fault in phase A, starting at 0.1
Note that (1) and (4) only apply for three phase fault, in second and clearing at 0.2 second is performed. The
case of unsymmetrical fault, the voltage magnitude and the instantaneous voltage waveform, RMS voltage and phase angle
phase angle jump in each phase depends on many factors such jump are shown in Fig.3.
as the point of fault initiation, the type of fault, the parameters
of devices in the system, etc.
C. Voltage sag duration
The duration of the voltage sag is defined as the time
between the sag starting and clearing. In the three phase
system, the voltage sag start when at least one of the RMS
drops below the threshold which is typically chosen as 90% of
the nominal voltage magnitude, the sag ends when all three
voltages have recovered above the threshold [6].
III. SIMULATION VOLTAGE SAG DUE TO SHORT CIRCUIT
A. Fault system model
The distribution voltage level used in the models is based
on the Vietnam grid code. Tab. 1 shows system parameters
used in the simulation. The simulation of voltage sags due to V=0.9p.u
short circuit was performed in EMTP software is shown in RMS sag
Fig.2. RMS sag ending
beginning at 0.217s
at 0.103s Duration
TABLE I. SYSTEM PARAMETERS

No. Components Details

1 Source 250 MVA,110kV, 50Hz, X/R=10.5

2 HV/MV Transformer 25 MVA, 110/22kV, DYn11

3 MV/LV Transformer 2 MVA, 22/0.4kV, DYn11

4 Line Overhead line, AC70, 5 km

5 Load 0.4 kV, 1.67 MW and 0.58 Mvar

1.04ms

- Voltage sag at Phase A - Voltage normal

Fig. 3. Single phase to ground fault. (a) Instantaneous waveform of three

phase; (b) RMS waveform; (c) phase angle jump in phase A.

Fig 3b, 3c show that the voltage in faulted phase A is

Fig. 2. One phase diagram of distribution system for simulation reduced by 50% during the voltage sag and slight overvoltage
appears at the end of the sag while the voltage in phase B and
It is assumed that a fault has happened at the end of 22 kV C just only reduced insignificantly. This phenomenon is
line. The measuring equipment is installed at the PCC. In the related to the type of the fault. Fig. 3 shows the phase angle
simulation model of voltage sags, voltage waveform magnitude jump in the faulted phase A as -18.720 whereas other un-faulted
is commonly presented in RMS waveform and normalized for phases are almost unaffected. The duration of voltage sags
better comparison. The transformer is modeled by (under 90% of the nominal voltage) in this case is 0.114
SATRANFO block, the RMS analysis is simulated by model second.
RMS1F and a user-specified library file (RMS1F.SUP) would
be created in ATPDraw.
C. Phase to phase fault 0.109 second; the phase angle jumps are -32.040 and +12.240
degrees respectively.
D. Two phases to ground fault and three phases fault

V=0.9p.u

- Phase A - Phase B - Phase C

Fig. 4. Phase to phase fault. (a) Instantaneous waveform of three phases; (b)
RMS waveform; (c) and (d) phase angle jump in phase B and phase C.

In this simulation, it is assumed that phase to phase ground

faults has appeared on phase B and C, from 0.1 to 0.2 second. Fig. 5. Two phases to ground fault. (a) Instantaneous waveform of three
The instantaneous voltage waveform, RMS voltage and phase phases; (b) RMS waveform; (c) and (d) phase angle jump in phase B and
angle jump are shown in Fig. 4. Observing in Fig. 4a, 4b the phase C.
phase voltage drop of magnitude Vb = 0.76 p.u , Vc = 0.50 p.u
occurs but phase voltage Va remains unchanged. The duration Similar results have been obtained for this system when
of voltage sags at phase B is 0.091 second, and in phase C is two phases to ground fault on phase B and C, and three phases
fault from 0.1 to 0.2 second is simulated. The instantaneous The transformer energizing model developed in
voltage waveform, RMS voltage and phase angle jump of these EMTP/ATP is shown in Fig. 7. It is used to simulate voltage
two cases are shown in Fig. 5 and 6. The simulation results sags caused by transformer inrush current and core saturation
obtained from this type of the fault are summarized in Tab. 2. during energizing. The model consists of 22 kV source,
25 MVA, 50 Hz connected to a three-phase breaker,
22/0.4 kV, 1 MVA, DYn11 saturable core transformer and 200
kW resistive and 150 kVar inductive load. The saturable core
transformer is modeled by BCTRAN block. The saturation
characteristic of the saturable transformer block is shown in
Fig. 8. The switch is set to open at initiating stage and close at
0.4 second, the voltage sags usually takes nearly 2 second to
rise back to its nominal voltage level.
1.1

1 l-i

.9

.8

.7

Fluxlinked [Wb-T]
.6

.5

.4

.3

.2

.1

20 40 60 80 100
I [A]

Fig. 8. The saturation characteristic of the saturable transformer

Fig. 6. Three phases fault. (a) Instantaneous waveform of three phases; (b)
RMS waveform.

TABLE II. SIMULATION RESULTS OF VOLTAGE SAG DUE TO SHORT

CIRCUIT

Voltage sag Phase Sag duration by Sag

Faults Phase magnitude angle jump RMS on phase duration
(p.u) (deg) (second) (second)
SLGF A 0.50 -18.72 0.114 0.114
B 0.76 -32.04 0.091
LLF 0.091
C 0.51 +12.24 0.109
B 0.54 -18.54 0.099
LLGF 0.099
C 0.49 -21.06 0.108
A 0.48 -25.90 0.108
TPF B 0.48 -12.60 0.109 0.108
C 0.48 -20.09 0.110
Fig. 9. Voltage sag RMS waveform caused by transformer energizing at 22
kV bus
IV. SIMULATION OF VOLTAGE SAG DUE TO TRANSFORMER
ENERGIZING Fig. 9 shows the RMS analysis of three phase voltages sag
caused by transformer energizing. It can be clearly seen that
this case results in unbalanced voltage sag and all voltages
gradually rise to its nominal voltage level. In this case, shallow
voltage sag of 2% is present, but it can be up to 15% of its
nominal magnitude. It is dependent on the source power rating,
the transformer power rating and the power consumption of the
load.
V. SIMULATION VOLTAGE SAG DUE TO INDUCTION MOTOR
STARTING
Motors have the undesirable effect of drawing several times
Fig. 7. Transformer energizing model in EMTP/ATP.
their full load current while starting. The induction motor
starting model ([7], [10]) is used to simulate voltage sag caused the power of transformer is much larger than power of load and
by starting a high power industry induction motor. The induction motor.
induction motor starting model developed in EMTP/ATP is
shown in Fig. 10. The induction motor is modeled by UM_3 VI. CONCLUSION
block [8]. The paper presents a technique to simulate the causes of
voltage sag by using ATP/EMTP software, including voltage
sags due various types of short circuits, transformer energizing
and induction motor starting. In these simulation models,
voltage sags characteristics is analyzed in term of the
magnitude, duration and phase angle jump. The simulation can
be useful to find the solutions in order to mitigate the voltage
sags. The results shown that EMTP/ATP is very flexible to
deal with various network configurations in steady as well as in
the transient states.
REFERENCES

[1] IEEE Std. 1159-1995, Recommended Practice for Monitoring Electric

Power Quality, 1995.
[2] Patne N R, Thakre K L, “Stochastic Estimation of voltage Sag Due to
Fig. 10. Induction motor starting ATPDraw model. Faults in the Power System by Using PSCAD/EMTDC Software as a
Tool for Simulation”, Journal of Electrical Power Quality and Utilisation
13, p59-63, 2007.
TABLE III. THE PARAMETERS OF INDUCTION MOTOR
[3] Filho J M C, Leborgne R C, Abreu J P G, Novaes E G C, Bollen M H J,
Poles Voltage Current R1 X1 Xm R2 X2 “Validation of Voltage Sag Simulation Tools: ATP and Short Circuit
(V) (A) (ȍ) (ȍ) (ȍ) (ȍ) (ȍ) Calculation Versus Field Measurements”, IEEE Transactions on Power
2 380 57 0.322 0.657 12.5 0.196 0.510 Delivery 23, p. 1472-1480, 2008.
[4] Hadi Saatdat, “Power system Analysis”. New York, McGraw Hill 1999,
ISBN0-07-561634-3.
[5] Meenu Kanwar, Komal Arora, Sawan Kumar Sharma, “Comparison of
Simulation Tools ATP-EMTP and MATLAB-Simulink for Time
Domain Power System Transient Studies”, International Journal of
Research Studies in Computer Science and Engineering (IJRSCSE)
Volume. 1, Issue 3, July 2014, PP 50-56, ISSN 2349-4840 (Print) &
ISSN 2349-4859 (Online).
[6] Math H.J. Bollen, “Understanding power quality problems”, A JOHN
WILEY and SONS, INC, 2000, ISBN 0-7803-4713-7.
[7] P.Pillay, S.M.A. Sabur, M.M. Haq, “A model for induction motor
aggregation for power system studies”, Electric Power Systems
Research 42, p. 225-228, 1997.
[8] Morris Lockwood, “Issues in Modelling Transformers, Machines and
their Controllers Using ATPDraw”, EEUG (European EMTP-ATP
Users Group) course 2010, 27-29 April, 2010.
[9] Lõszlú prikler, Hans Kristian Hoidalen, “ATPDRAW”, version 5.6,
2009.
[10] Rodney H.G. Tan and Vigna K. Ramachandaramurthy, ”A
Comprehensive Modeling and Simulation of Power Quality
Disturbances Using MATLAB/SIMULINK”, available at
http://www.intechopen.com/books/power-quality-issues-in-distributed-
generation/a-comprehensive-modeling-and-simulation-of-power-quality-
disturbances-using-matlab-simulink
[11] Suresh Kamble, Chandrashekhar Thorat, “Voltage Sag Characterization
in a Distribution Systems: A Case Study”, Journal of Power and Energy
Engineering, 2014, 2, 546-553, Published Online April 2014 in SciRes.
http://www.scirp.org/journal/jpee.
Fig. 11. Voltage sag RMS waveform of 22 kV feeder voltage and 0.4 kV [12] Shailesh M. Deshmukh, Bharti Dewani, S. P. Gawande, “A review of
load voltage caused by starting of 50 hp induction motor at 0.4 kV feeder. Power Quality Problems-Voltage Sags for Different Faults”,
International Journal of Scientific Engineering and Technology, Volume
Fig. 11 shows a three-phase voltage sags of RMS No.2, Issue No.5, pp. 392-397, 1 May. 2013, ISSN : 2277-1581
waveform at 22 kV and 0.4 kV due to an induction motor
starting upon closing of motor starting contactor at 0.2 second.
The sag magnitude is dependent on the induction motor and
transformer power rating. In this case, it is approximately 6%.
The voltage sag at 22 kV feeder not noticeable (<1%) because