High voltage transmission lines are designed to transmit electrical energy from one place to other place. The faults occur in power system High voltage transmission lines Above 80%. Faults or abnormal condition occur due to external causes or internal failures in High voltage power transmission lines. Types of fault and location of fault identification is very essential for the system. It reduces Outage time. Fault identification is a difficult task. There is no standard method for finding fault location. There are various reason behind occurrences of fault like as insulation failure, breakdown of insulator, faulty tripping of a circuit breaker. In practical fault can occur at any location and at any instant of time. The fault resistance can vary from low resistance (few ohms) to high resistance as (100Ω), which influence the transient characteristics of the voltage waveforms irrespective of the type of fault. This change of resistance makes fault identification is more difficult. Our main objective is to identify the fault and possible location of fault as early as possible. An accurate digital model of a test system becomes necessary to perform the simulation and result purpose, as practical experiment is difficult over here. Here we used EMTP Software as it gives the accurate response at transient condition as compared to real value data.
Keywords:- High voltage transmission line, fault, S-transformation, EMTP/ATP Software.
1. INTRODUCTION
High voltage transmission line carries a vital role in the power system network. It helps to deliver power from one place to other place. Therefore, the reliability and continuity of the power supply is one of the major challenges faced, because High voltage transmission lines are most prone to faults. Whenever a fault occurs in High voltage transmission line it is necessary to locate and identify the fault as early as possible such that it will not harm the whole power system network. Faults in High voltage transmission lines occur because these are exposed to severe environmental conditions like falling of a tree branch, storm, lightning strokes etc. Insulation failure and broken insulators also cause faults in the High voltage transmission lines. In an electric power system, a fault is any abnormal flow of electric current. Fault diagnosis involves three main tasks: fault detection, classification, and location of fault. In three-phase systems, at fault condition involve one or more phases and ground, or may occur only between phases. The balanced faults are three phase(L-L-L) and three phase to ground(LLL-G)short circuits. Most of the faults occurred are of unbalanced in nature. Single-line to ground (L-G), line to line (L-L) and double-line to ground (LL-G) faults. Different researchers have developed a number of fault classification techniques. Wavelet multi resolution analysis along with adaptive neuro fuzzy [1] interference system is a modern technique for fault classification and identification. Artificial neural network [2] approach has been applied to fault classification and location in High voltage transmission lines. Employing wavelet[3] transform for detecting arrival time of waves to terminals for fault location[4] in three terminal lines is proposed[5].One terminal voltage and current signals and then by using Radial[6] Basis Artificial Neural Network (RBFNN) algorithm has found the fault location. By voltage[7] and current from one terminal to locate fault. By Fuzzy [8] logic for fault[9] distinction.. [10] S.M. Brahma, uses voltage measurement from all terminals to estimate the fault location in a Multi- Terminal High voltage transmission line. In this paper, Our aim at using S-transformation to classify and find fault location in High voltage transmission line. The Different fault conditions are simulated using electromagnetic transient programme (EMTP) software.
2. S-TRANSFORM THEORY:
The S-transform is a time-frequency [12] analysis technique with variable window width inversely varies with frequency. It is extension of wavelet transform. S-transform has a frequency dependent resolution of time-frequency domain and entirely refer to local phase information. S transform is a special case of STFT with Gaussian window function. If the window of S transform is wider in time domain, S transform can provide better frequency resolution for lower frequency. While the window is narrower, it can provide better time resolution for higher frequency. S-transform is convertible both in forward and inverse from time domain to frequency domain. The generalized S-transform is defined as
(1)
Where w is the window of S-transform, p denotes the set of parameters that determines the shape and property of w. is the position of w on time axis.
(2)
Alternative expression of equation (1) obtains with the Fourier transform of w and h can be given as:
(3)
Where
(4)
. (5)
H and W is the Fourier transform of h and w respectively. The variable α and f has the same units.
In discrete S-transform h(t) is expressed in terms of h [kT]. k is the discrete time series of the signal h(t) and it varies from 0 to N-1. Discrete Fourier transform of sampled time series can be written as:
(6)
Where discrete time k=t/T=0,1,..........,N-1. N=no of sample, Discrete frequency n= f*N*T, τ=j*T. T=sampling interval.
The S-transform of discrete time series is expressed as:
Where j,m and n=0,1,............,N-1.
Small value of α improves time resolution but reduces frequency resolution. Large value of α reduces time resolution but improves frequency resolution.
3. SYSTEM SIMULATION:
200kmmm
Source(Es) Er
Fault
Fig.1: Transmission line model
132 KV, 3ph, 200km line is taken up for study. Alternative Transient Program/Electro Magnetic Transient Program(ATP/EMTP) software is used for fault simulation. The sampling rate is used 1.0 kHz at 50 Hz frequency and 20 sample per cycle. The parameters used here is as follows: Resistance 5 ohm/km, Inductance 0.99584 mH/km, Capacitance 0.012416 μF/km. Tower footing resistance is 100 ohm. Fault resistance 18 ohm. The transient fault types are single line to ground
(SLG), line to line (LL), and 3-phase fault (PF). Fault location was taken in between 20km to
200km of the High voltage transmission line length from the source of fault.The signal is processed by using discreteS-transform. By discrete S-transform S-matrix is formed, which is used to calculate change of energy. Change in energy is calculated for fault detection and phasors are estimated for impedance calculation.
Fig.2: fault voltage and current signal for single line fault generated by EMTP Software
4. DISTURBED PHASE IDENTIFICATION:
Faulty signal for current and voltage wave are taken from the relay coil. The data of current and voltage waveform are taken one sample of the inception of fault. Then the S-transform is computed for current and voltage wave in different fault condition. Which gives the energy content of that faulted signal. This procedure is done again for pre-fault signal. Now the energy difference between faulted signal and pre-fault signal indicates the faulted phase and un-faulted phase clearly.
Formula: Energy change=Ef-E n={abs(Sf)}2-{abs(Sn)}2 (8)
The energy change of faulted and un-faulted signal with different fault resistance and different fault inception angle is shown in the above table. More energy change is occurred in faulted phase than un-faulted one. For B-G fault with fault resistance Rf=100ohm, and fault inception angle δ=45,at 10% of the transmission line ie 20km energy change in current signal are 0.4142, 55.7754 and 2.7181 for A, B, C phases respectively. Energy change in voltage signal are 0.264, 14.44, 1.29 for A, B and C phases respectively. The higher value of energy change for the respective phases identifies the faulty phase from the un-faulted phase.
Table-I: Spectral Energy change for different fault condition
Faults
Voltage
Current
A
B
C
A
B
C
A-G(Rf=100 Ω,δ=45°,at 10%ie 20km)
33.27
0.15
0.132
50.6
14.68
1.85
B-G(Rf=100 Ω,δ=45°,at 10%ie 20km)
0.264
14.44
1.29
0.4142
55.7754
2.7181
C-G(Rf=18Ω, δ=0,at 45% ie 90km)
0.345
0.644
14.75
1.3765
16.3591
563.022
AB_G(Rf=30Ω,δ=60°, at 30% ie 60km)
26.01
38.15
2.19
57.35
76.35
10.75
BC_G(Rf=30Ω,δ=0°, at 45% ie 90km)
3.66
24.87
15.023
1.6997
160.8322
545.1532
AC(Rf=30Ω,δ=0°, at 45% ie 90km)
12.6
1.18
15.04
16.614
0.246
57.806
ABC_G(Rf=18Ω,δ=0°, at 52.5% ie 105km)
16.02
43.42
31.82
29.22
18.01
88.9
ABC_G(Rf=18Ω,δ=0°, at 37.5% ie 75km)
11.29
15.56
6.39
11.38
11.48
41.86
ABC(Rf=150Ω,δ=30°, at 70% ie 140km)
61.127
2193.13
472.413
30.19
24.27
31.48
Table -II: Per Unit bassed Spectral Energy change for different fault condition
Faults
Voltage
Current
A
B
C
A
B
C
B-G(Rf=100 Ω,δ=45°,at 10%ie 20km)
0.0182
1.000
0.0893
0.0074
1.000
0.0487
A-G(Rf=100 Ω,δ=45°,at 10%ie 20km)
1.000
0.0045
0.0039
1.000
0.2901
0.0365
B-G(Rf=100Ω,δ=45°,at 12.5%ie25km)
0.0212
1.000
0.0097
ABC_G(Rf=18Ω,δ=0°, at 52.5% ie 105km)
0.3689
1.000
0.7328
0.3286
0.2025
1.000
ABC_G(Rf=18Ω,δ=0°, at 37.5% ie 75km)
0.7256
1.000
0.4107
0.2718
0.2742
1.000
ABC_G(Rf=150Ω,δ=30°, at 70% ie 140km)
0.9590
0.7709
1.000
AB_G(Rf=30Ω,δ=60°, at 30% ie 60km)
0.6817
1.000
0.0574
0.7511
1.000
0.1407
The amplitude and phase of faulted signal are also calculated from S-transform matrix. First calculate the amplitude by given formula in MATLAB
Amplitude=max(abs(S));
S=S-transform Matrix, abs=absolute value, max=maximum value
Now we plot a graph between Amplitude vs frequency. The frequency at which Amplitude is maximum known as faulted current frequency. The instantaneous phase of the signal is found out at the exact frequency voice. S-tranform matrix is a complex one. So the phase is calculated by the MATLAB command
Phase=atan(imag(S)/real(S));
5. FAULT LOCATION:
After determination of fault type from energy change table we have to calculate the location of fault. Here we used polynomial curve fitting to find out the fault location. The curve fitting is done on an index.
Index=│Ev│/│Ei│
Where Spectral energy change of faulted voltage and faulted current signals are Ev and Ei respectively. The index is calculated at different fault locations. It is prepare with different fault resistance and different fault inception angle.
The polynomial is used here
Y=A0+A1x+A2x2+A3x3+A4x4+A5X5
Y is the fault location and x represents the index value. Fault location is find out by above 5th degree of polynomial.
Calculation of error:-
SINGLE LINE TO GROUND FAULT
DISTANCE
ACTUAL FAULT DISTANCE (Km)
CALCULATED FAULT DISTANCE (Km)
PERCENTAGE OF ERROR
25% of Line
50
52.68
5.36
35% of Line
70
68.8845
-1.594
50% of Line
100
98.911
-1.089
52.5% of Line
105
104.3397
-0.629
65% of Line
130
130.7296
0.561
70% of Line
140
137.8561
-1.531
87.5% of Line
160
162.4206
1.513
AB TO GROUND FAULT
DISTANCE
ACTUAL FAULT DISTANCE (Km)
CALCULATED FAULT DISTANCE (Km)
PERCENTAGE OF ERROR
30% of Line
60
59.4919
-0.847
35% of Line
70
69.1039
-1.280
50% of Line
100
101.2491
1.249
70% of Line
140
140.2184
0.156
80% of Line
160
159.851
-0.093
87.5% of Line
175
174.4586
-0.309
ABC TO GROUND FAULT
DISTANCE
ACTUAL FAULT DISTANCE (Km)
CALCULATED FAULT DISTANCE (Km)
PERCENTAGE OF ERROR
20% of Line
40
40.9282
2.320
35% of Line
70
70.7508
1.073
52.5% of Line
105
105.6879
0.655
65% of Line
130
132.3152
1.781
70% of Line
140
140.5453
0.389
80% of Line
160
164.1032
2.564
87.5% of Line
175
174.7903
-0.120
6. CONCLUSION
In this paper we are using computer with PLC based logic for
