Design of Simulink Model for Three

Phase Transmission Line Fault

Detection
Purujit Chaturvedi1 Ravi Raj2
Department of Electrical Engineering, IKG Punjab Technical University Jalandhar, Punjab

Abstract - After power generation, power transmission is one of the most important issues in electrical
engineering. Transmission line faults are a widespread and serious concern in this stream. This study
describes a technique for detecting the location of various defects on transmission lines so that protection
schemes can operate quickly and reliably. To generate the results, the simulation was created in MATLAB.
Transient voltage and current are made up of fundamental components. The MATLAB software is used to
model various operating and problem scenarios. single phase to ground fault, line to line fault, double line to
ground, and three-phase to ground faults on high-voltage transmission lines short circuit in phase Variations
in fault resistance (Rf) and distance to fault (Lf) have had a wide range of effects on voltage and current and
its relationship to the system's impedance, which generates the logic for fault detection, categorization, and
localization.

Keywords: Transmission Line faults, Transmission Line Protection, Detecting and Locating
faults in overhead transmission lines, Fault Analysis in transmission lines.

1. Introduction
Fault location and distance estimation are critical issues in power system engineering in order to
swiftly clear faults and restore power supply with minimal interruption. [1] This is required for
the safe operation of power equipment and customer satisfaction. In a three-phase transmission
line, there are 11 probable faults: AG, BG, CG, AB, AC, BC, ABG, ACG, BCG, ABC, and
ABCG.
There are five main sorts of flaws that can be classified as follows:
1. LG faults – AG, BG, CG
2. LL faults – AB, AC, BC
3. LLG faults – ABG, ACG, BCG
4. LLL faults – ABC
5. LLLG faults – ABCG
 Figure 1: Simple Power System Network
Faults can be easily identified by a sudden drop in line impedance due to high current during the
fault.
The next issue is its classification, which refers to the type of fault that has happened. The LLL
and LLLG faults are symmetrical faults that are indistinguishable among these five. During the
fault, the voltage and current values quickly shift, and phase imbalance ensues. In each phase,
the approach necessitates a constant inspection of the line impedance readings.[2] The phase
with a sudden decrease in impedance indicates a defective phase.A zero sequence analyzer can
tell if the ground is implicated in the fault or not.
In the event of a ground fault, current flows in a zero sequence. This work provides a fault
detecting and locating technique based on line impedance monitoring. MATLAB's power system
BLOCKSET is used to simulate a 220KV, 200km, 50Hz transmission line.

2. Basic idea of the method used

When a defect occurs in the system, the impedance drops dramatically due to the current's short
circuit path. With the help of logic-controlled devices, this change in value is detected, and the
problematic phases are identified. Because the change in value is considerable and happens only
in the phases where the fault occurs, it is easy to identify problematic phases, and those phases
can be isolated from the system using a circuit breaker. Figure 2 shows the relay system
configuration for fault analysis. The basic component of each phase's voltage and current is
continually sent to the relay, which calculates the phase's impedance value.

The power system BLOCKSET of MATLAB is used to simulate a 220KV, 50Hz three phase
transmission line. All conceivable faults are simulated with various fault resistance values and
locations, yielding the set point value for current and impedance to detect the fault and identify
the phases. We employ a Zero Sequence Analyzer to determine whether or not ground is
involved in the fault, and it only delivers a decent result when ground is involved in the fault.
Figure 3 shows the Simulink model that was employed.[3] The three-phase fault block allows for
the simulation of all potential faults from any inception angle. The voltage and current block is a
subsystem block that includes all of the measuring equipment needed for relay operation.
 Figure 2: Single line diagram of a three phase transmission line under study.

During a fault state, the impedance lowers from thousands to hundreds, while the current rises
from hundreds to thousands, depending on the fault resistance and distance to the fault. These
numbers alter as the load changes. High load currents can cause issues for single-ended fault
location methods, hence they must be considered. For proper fault and overload discrimination,
the right set point must be chosen.

Figure 3: Power system model in Simulink under study

3. Algorithm to detect and classify fault

Figure 4 shows graphs of fundamental current levels in all three phases, as well as a Zero
sequence analyzer simulated during a fault. In the zero sequence analyzer, we detect a large peak
 with a value of 150 A. (under normal conditions the value is near to zero). This high score
indicates that Ground is a part of the fault. Because the current in phases B and C is below
normal limits (below 160A – 190A), they are not engaged in the fault. In Phase A, the current
rises dramatically, resulting in a fault. As a result, the fault type is AG.

As a result, any phase in which a large current more than 200 A flows is designated as defective.
If the zero sequence current value is greater than 35 A, it is determined whether or not the
ground is defective. The set points are for a system with a 160 A nominal current in each phase.
[4] To distinguish between an overload state and a fault, a 40 A gap is maintained. These set
points can be altered in the same way for any other system. As the fault is classified by the relay,
trip signals can be delivered from the relay to the phases implicated in the fault.

3.1 Algorithm to locate the fault

The next task is to locate the defect, for which we took some notes. Line impedance varies
not only with fault resistance but also with fault location. Figure 5 shows the figure obtained
using Simulink by producing LG and LLG faults with varying fault resistance and position.
[5] The figures produced from simulation show that line impedance increases as fault location
and fault resistance increase. We have a database of fault line impedance measurements at
various fault locations and fault resistance levels. If, for example, a problem occurs that is
characterized as an LG fault, and the line impedance measured by the relay is 124Ω
(example from the model simulated, value can be referred from the plot) it is obvious that
fault resistance is of the resistance is either 100 ohms or 50 ohms. To improve accuracy,
measurements will be taken from the remote end, and the position will be matched up with
the line impedance value, resulting in the location of the fault. This can be accomplished
using any soft computing technique, such as fuzzy logic or artificial neural networks (ANNs)
(Artificial Neural Network).
 (a) Zero Sequence Current (c) Current in Phase B

(b) Current in phase A (d) Current in Phase C

Figure 4 : Plot of fundamental values of current in all three phase and zero sequence analyzer
(from SIMULINK)
 (a)
(a)
(a)
(a)
(a)
(a)
(a)
Fault type: LG (b) Fault type: LL

Figure 5: Plot for change of line impedance with change in fault location with different fault
resistance (obtained from SIMULINK) for (a) LLG fault and (b) LL fault.

4. Test Result of the Fault Detector and Classifier

The line impedance based fault detector and classifier is tested for different fault types
originating at different locations and at varied inception angles after the logic has been set up.
The fault classifier's answer is shown in Table I.
For example,[5] fault AG is formed with a fault resistance of 100 ohms at an inception time of
0.04sec (at a fault inception angle of 0 degrees). The currents in the corresponding phases will
alter after a fault occurs. Figure 4 depicts a plot of the fundamental value of currents in all phases
under this fault scenario. The error occurs at sample number 49 (49 X 0.000833sec =
0.040817sec), where 0.000833sec is the sampling time. Because the set point for the given
system is 200 A to consider fault in phase so from fig. 4(b) fault is detected at sample number 54
(T= 54 X 0.000833 = 0.044982sec).

The time of operation of line impedance based fault detector and classifier can be calculated as
follows:
Fault occurred at sample no. 49
Fault inception time = 49 X 0.000833 = 0.040817sec.
Fault detected at sample no. 54
Fault detected at time= 54 X 0.000833 = 0.044982sec.
Time of operation = Fault detection time – Fault inception
time = 0.044982-0.040817= 0.004165 sec.

The frequency of the system is 50 hertz. It takes 0.02 seconds to complete one cycle. It is
obvious from the preceding estimate that the time. It takes about a quarter of the time for the
relay to work. The system's time span (0.02 sec).

The testing data set was constructed using various failure scenarios, and Table I shows the results
of the line impedance based Detector and Classifier tests. In the corresponding faulty phases, the
output of the fault detector and classifier should be zero for a healthy condition (no fault
condition) and one for a faulty condition (fault condition). It is obvious from Table I that when a
defect occurs, the output of the associated phases increases.

Table 1 : Data for types of fault and fault location

Fault Type Fault Location Fault Fault Phase Phase Phase Ground
(Distance from inception Resistance A B C G
the relay in km angle (in in Ohms
degree)

No fault - - - 0 0 0 0

AG 100 0 100 1 0 0 1

AB 50 90 25 1 1 0 0

ACG 150 90 100 1 0 1 1

BG 20 50 25 0 1 0 1

ABC 50 50 100 1 1 1 0

The approach's speed and selectivity are quite robust, according to test results, and give adequate
performance for a three-phase transmission line.

5. Conclusion

The most common pre-known line parameters are compared to the impedance measured in the
case of a fault in impedance-based fault detection, classification, and location procedures.
Without using any of the bridge approaches, this method detects, classifies, and offers probable
fault location using a predefined set of variables and any soft computing technique.
References

[1] Y.G. Paithankar and S.R. Bhide, Fundamentals of Power System Protection, Prentice-Hall of
India Private Limited, New Delhi, 2003
[2] Mahmoud S. Awad, “On-line Determination of the Fault Location in a Transmission Line”,
Int. J. Emerg. Sci., 2(2), 210-221, June 2012 ISSN: 2222-4254, pp. 210 – 221.
[3] “Fault location on distribution systems: An update on EPRI and DOE research”, IEEE
Distribution Subcommittee presentation, Orlando, FL, Jan. 10, 2007.
[4] Satya Prakash and S.C.Gupta, “Fuzzy Logic Based Trained Fault Locating Mechanism in
Power Distribution Network”, ISSN 2250-2459, Volume 2, Issue 7, July 2012, pp. 129 – 135.
[5] C.L. Wadhwa, Electrical Power Systems, 4th ed., New Age International (P) Ltd., 2005.
[6] K. Erenturk, “A new digital protective relay based on fuzzy logic and value estimation”,
Iranian Journal of Science & Technology, Transaction A, Vol. 29, No. A2 Printed in The Islamic
Republic of Iran, 2005, pp. 267 – 276.