A COMPARATIVE STUDY: FAULT DETECTION METHOD ON OVERHEAD TRANSMISSION LINE

Transcription

1 Volume 118 No , ISSN: (on-line version) url: ijpam.eu A COMPARATIVE STUDY: FAULT DETECTION METHOD ON OVERHEAD TRANSMISSION LINE 1 M.Nandhini, 2 M.Manju, 3 S.Srinivasan 1,2 PG Scholar, 3 AP/Department of EEE, 1,2,3 Vivekanandha College of Engineering for Women,Elayampalayam, Namakkal, Tamilnadu, India 1 nandhinimani95@gmail.com, 2 manjumay13@gmail.com 3 srivaas131985@gmail.com Abstract: In this paper, the authors have carried out fault investigation in the transmission line in power system using two different techniques: Travelling wave method, wavelet multi-resolution technique. This method based on the measurement and analysis of voltage and current. The Line to ground fault is the most common fault (LG) in power system percent of the problems are the line to ground fault. Many kinds of researchers are even going on to enhance the techniques to solve these problems. Protection of transmission line is very important because it is a vital component between the generating stations and end users. A two-terminal 300kV transmission line fed by two generators model is simulated with MATLAB for all the two different techniques. Keywords: Single line to ground fault, travelling wave method, wavelet multi-resolution technique. 1. Introduction Most of the faults are occur in the power systems are unsymmetrical faults. Like single line to ground fault, double line to ground fault, three-phase fault and line to line fault. 75% of the faults are the single line to ground fault. Remaining three faults occur overhead transmission line when compared to the single line to ground faults these faults occur very low. The travelling wave is hard to separate from interference noise as a high-frequency signal. Travelling waves are the best choice to analyze the transient signal. Travelling waves are associated with noise.fast detecting, locating and repairing of these faults are critical. These calculations are used to define the most suitable technique for the detection of the Travelling wave method, wavelet multi-resolution technique under examination. The concept of travelling wave theory is useful for transmission line fault location. The connection between the bands of travelling waves, the terminal states of transmission lines and the fault distance was explored [1-2]. The high-frequency signal has been exploited in travelling wave based fault identification. These signals are filtered from the measured signal [3]. Power transmission systems use protection relays to detect and clear different types of faults as soon as possible with total selectivity in order to minimize damages in the power elements that form the electrical power network as well as to reduce the instability in the network caused by such faults. Some transmission lines have two different parts, one is the underground cable side and the other one is the overhead side [4]. Different transformations can be used to observe the frequency contents of the fault signal. To observe the frequency contents of the fault signal. Fourier Transform (FT) is the transform technique easy to apply for fault location algorithm [5-6]. The frequency evidence of the signal is clearly given by Fourier Transform, but it does provide any details about the time domain specification. The data provided by the integral links to all-time instances since for all time intervals the integration is done. Therefore, we can conclude that no matter where in time the frequency appears, it will affect the outcome of the integration in the same way only. This is why Fourier Transform is not appropriate for non-stationary signals. Therefore, Even though the Fourier Transform in the frequency-domain has worthy results but in the time domain, it has very poor results[7]. Under travelling wave method, we have single-ended fault location algorithm and double-ended fault location algorithm. In that paper single-ended algorithm used. The travelling of the first wave fears from the fault point to terminal. The arrival of wave after reflecting back from fault point is always proportional to fault distance [8-9]. This paper presents a new protection method for single-phase ground faults detection in cables used in overhead cable lines in power transmission systems. If the magnitude differences are less than a threshold M, the fault is identified as fault outside the cable line side. But if the magnitude differences are higher than such threshold M, the fault is identified as fault inside the cable line side[10]. In this paper, describes the Wavelet multiresolution technique, and travelling wave method which can be applied to a Transmission Line Fault 961

2 Analysis. A two-terminal transmission line fed by two generators is used for the fault investigation. For this, different realistic fault cases at different locations were considered. For each assumption, adequate details and explanations are given for more clarity & better understanding. For the fault investigation, a two terminal 300kV transmission line fed by two generators is used for two different techniques. 2. Wavelet Multi-Resolution Technique Mother wavelet used as a characteristic function in an analysis to vary the multi-resolution techniques. This prototype function will sharply decay in an oscillatory manner and has a zero mean, i.e., it suddenly drops to zero on each side of its central path. Precisely, the CWT (Continuous Wavelet Transform) of a given signal (t) with respect to a characteristic function ȟ (t) is generally defined as in equation 1, Where, ȟ(t) is the mother wavelet, and the other wavelets ȟp,q (t) = ȟ are dilated and translated versions of the mother wavelet ȟ(t). The constant p is called dilation or scale parameter, which is inversely proportional to frequency and the constant q is called translation parameter, which refers to the position of the window. The Discrete Wavelet Transform (DWT) is the digitally implementable equivalent of the Continuous Wavelet Transform (CWT). Accurately, the Discrete Wavelet Transform (DWT) of a given signal Ȥ(t) with respect to the characteristic function ȟ(t) is generally defined as in equation 2, Where ȟ(t) is the mother wavelet. The constants p and q dilation and translation parameters are functions of an integer parameter m, giving rise to several daughter wavelets. The k is an integer variable that denotes to an exact sample number in an input signal. The mother wavelet that has a high correlation with the high frequency travelling wave signals is considered as the best wavelet for fault detection. Daubechies wavelet is one of such mother wavelets appropriate for a wide variety of power system applications. It is the due to its inbuilt characteristics. Preferably, it is appropriate for the short duration, sensing low amplitude, fast decaying and oscillating type of signals, usually that seen in real time power systems. For detecting and localizing various types of faults the choice of analyzing wavelets plays a substantial part in it. However, the choice of a suitable mother wavelet without knowing the type of transient disturbance is a difficult task. Therefore, as an alternative for producing algorithms to select suitable wavelets, we can use one general type of mother wavelet for detecting and focusing on all types of disturbances without considering its duration. However, higher scale signal decomposition is essential in wavelet so the mother wavelet is at most localized in time and it can oscillate most rapidly within a very short period at the lowest scale that scales one. The analyzing wavelets become less localized in time and oscillate less as the wavelet goes to higher scales. Dilation nature of the wavelet transforms analysis in multi-resolution. At lower scales, to be identified fast and short transient disturbances, whereas slow and long transient disturbances, will be identified at higher scales due to higher scale signal decomposition. Therefore, we can distinguish fast as well as slow transients with a single type of mother wavelet. In this technique to be identified between different detail coefficients of Daub4 wavelets, which refers to different decomposition levels, only the summation of the sixth-level detail coefficients (d6) is considered. Wavelet MRA can investigate a signal with mixed frequency bands containing different resolutions. It is done by decomposing the signal into the highfrequency component, which is known as its approximation (a1h), and low- frequency component, which is known as its detail (d1l). The high-frequency component (a1h) can be decomposed further into another approximation (a2h) and another detail (d2l), and this technique is repeated. Therefore, we can conclude that the approximation of the signal is achieved by filtering the signal through a low-pass filter and decomposition is achieved by filtering the signal through a high-pass filter. So, consecutive filtering of the time-domain signal will give the required decomposition. The consecutive stages of decomposition are known as the levels. As pointed earlier, for detecting and localizing various types of faults the choice of analyzing wavelets plays a substantial part in it. However, the choice of a suitable mother wavelet without knowing the type of transient disturbance is a difficult task. Therefore, the alternative wavelet for producing algorithms to select suitable wavelets, we can use one general type of mother wavelet for detecting and focusing on all types of disturbances without considering its duration. However, higher scale signal decomposition is essential for the mother wavelet is utmostly localized in time and oscillates most rapidly within a very short period at the lowest scale that is scale one. The analyzing wavelets become less localized in time and oscillate less as the wavelet goes to higher scales. 3. Travelling Wave Method In the travelling wave method presents on the double terminal methods of travelling wave using wavelet. It will generate both forward and backward TWs signals 962

3 propagating away from the disturbance point towards both bus bars. The initial values of the waves are dependent on several factors such as fault position, fault path resistance, fault inception angle, and several types of fault, etc. Furthermore, these signals will be reflected and refracted at the points of discontinuity. In this points of discontinuity fault point and bus bars, until they are attenuated to a negligible value. The basic principle of this method can be well explained using Bewley lattice diagram Wavelet transform has the much better resolution to locate a transient event in time domain such as Fourier transform method. The waves reach both the bus bars after a certain time delay which is actually equal to the time. Further, it can be required by the wave to travel the distance with speed of light. Upon reaching the bus bar, some part of the wave is transmitted to the other medium and rest is reflected back to the same medium. The same pattern is repeated when these newly generated waves represented subsequent arrows reaches again to the bus bars. The generated transient signals consisting of different high-frequency components with information on the type of fault can be used to locate the fault. Since the main objective of this investigation is to classify the type of fault and to locate the fault. The distance to the fault location from the starting point is calculated based on the following dependence: (3) D Distance to fault location (m) t1 time in which the wave is generated as a result of switching (s) t2 time in which the reflected wave reaches starting point (s) v Wave propagation velocity (m/s) In this method may be used to check whether in the electrical length of the operating line corresponds to the line length measured using another method. Such a procedure is based on switching off the line breaker and then measuring the time in which the reflected wave returns to the locator. 4. System Description and Simulation In this paper, detecting and classifying faults on overhead the power transmission lines. In Wavelet multi-resolution technique and travelling wave method has been used. A three-phase model has been considered for this simulation study. For the simulation study, a fault analysis in overhead transmission line has been considered as shown in figure 1. Simulation model consists of two generators(g1, G2) and two distributed parameter line 1 and distributed parameter lines 2, three- phase V-I measurement 1, three-phase voltage measurement 2, and then fault block. The length of transmission line is 300km. A phase load is connected to load busb2. A three-phase load is connected to the load bus B2. The fault can be identified and distinguished from a power swing by measuring the energy level of 6th level and 11th level wavelet. In the 11th wavelet to provide the information about power swing to analysis the overhead transmission. Figure 1. Fault analysis in overhead transmission line 963

4 11th level wavelet exceeds the threshold value in the presence of power swing is identified.6th level wavelet to give detailed information about fault classification. The current phases are a,b,c respectively Sa, Sb, Sc to be obtained. Fault can occur in line when the summation of Sa, Sb, Sc is equal to zero. Most of the fault is the occur in the power system are the line to ground fault similarly 65 to 70 percent of the fault can occur in this type. In case of an L G fault occur in the transmission line means the absolute value of summation value used in one phase. In (Sb) is always much greater than the absolute value of summation of the other two phases (Sa & Sc). In this case, the absolute value of the other phases, namely Sa & Sc will be nearly equal and will be almost equal to zero. For example, double line to ground fault occurs in the transmission line, the absolute value of any two sixth level summations (Sb & Sc) is always much higher than the absolute value of summation of the other phase namely Sa. 5. Result and Discussion In this section to be described line to a ground fault can be analyzed. The various types of fault occur at the different distance on power transmission line of 300km. In an overhead transmission line, most of the fault occurred and line to ground fault is common the fault. The line to ground fault is analyzed for on phase B. Figure 2 & 3 represents the voltage waveform and current waveform in case line to ground fault (L-G). In all above cases, in order to compare the results. The fault is created at 90 Km from the bus bar, B1 and the fault creation time is 1.9 sec and the waveforms appear at sec. this is due to the travelling time taken by the fault to appear at relay point. so, difference in time is [ =0.004s]. Figure 2. voltage waveform of line to ground fault (L-G) on Phase B In the above Voltage, waveforms show the output of line to ground fault(l-g). In this above voltage waveform, three phases must be present and the affected phases are presented in the simulation result. In before fault, all three phases are which is normally working without fault is exist in the simulation. In this waveform shows the line to ground on phase B. Figure 3. Current wave of line to ground fault (L-G) on Phase B 964

5 In the above Current, waveforms show the output of line to ground fault(l-g). In this affected phases are presented in the simulation result. In before fault, all three phases are which is normally working without fault is exist in the simulation. By travelling wave method, the fault location is estimated by multiplying the difference in time with the velocity of wave propagation, which is about 99km. By the wavelet multi-resolution technique the fault is described in section II. The corresponding result is presented and compares with travelling wave method and wavelet multi-resolution techniques in table 1. Table 1. Comparative results for line to ground fault Fault distance (Km) Wavelet multiresolution technique Calculated Error Travelling wave method Calculated Error (Km) % (Km) % When we compared to other technique, the fault detection using travelling wave method is faster them analyzing travelling wave, which is fluctuating signal. As a high-frequency signal, the travelling wave is hard to separate from interference noise. A number of signals processing techniques are examined in this techniques used for identifying the fault location in the overhead transmission line. The multi-resolution technique provides a better result in fault calculation compared to another method. Travelling wave method provides better outcomes in the situation of fault classification. In addition, when more than one phase is involved in a fault, the travelling method provides better outcomes in estimating the fault distances. However calculating fault distance in case of line to ground faults (L-G), there more common, the multi-resolution technique does not deliver the clear result. Even though the summation of wavelet multiresolution analysis coefficients contains all information concerning the frequency components connected with any fault in a Transmission line, they are in raw form. The error is calculated in the fault distance and the error is less in the situation when one phase is affected by a fault. Therefore, we have the relevant data from the available data, so that the results will be less accurate when compared to travelling wave method. we can represent such acquaintance, then the overall uncertainty will be minimized. So, soft-computing techniques are used efficiently to take the information from available raw data for the location of faults in power transmission lines. Finally, travelling wave method provides accurate result compared to other techniques. 6. Conclusion A comparative study establishes that travelling wave technique establish that distinguishes a fault from a power swing much better than the other techniques. The travelling wave method for detection and classification of transmission line faults. Travelling wave method is very effective by using the wavelet travelling wave multi-resolution technique, to take the different harmonics generated in power swings and fault. Simulation results show the logic is fully deterministic, easy to understand, and also the classifier operation is fast and reliable. The techniques were tested in MATLAB using data generated by executing different faults. References [1] C. Pothisarn and A. Ngaopitakkul, Wavelet Transform and Fuzzy Logic Algorithm for Fault Location on Double circuit Transmission Line, 16th ICEE International, Busan, Korea, July 11-14, [2] Atthapol Ngaopitakkul et al., Combination of Discrete Wavelet Transform and Probabilistic Neural Network algorithm for detecting Fault Location on Transmission Systems, 14th ICIC International, Vol. 7, No. 4, pp , April [3] Du Lin, Pang Jun, Sima wenxia, Tang Jun, Zhou Jun Fault location for transmission line based on Travelling waves using correlation analysis method 2008 international conference on high voltage engineering and application, Chongqing, China, November 9-13, [4] N. Tleis. Power Systems Modelling and Fault Analysis, Newnes. Oxford, [5] Das, S. SANTOSO, S. Gaikwad, A. Patel, "Impedance-based faultlocation in transmission networks: theory and application," Access, IEEE, vol.2, no., pp.537, 557, May [6] Aritra Dasgupta, Sudipta Nath and Arabinda Das, Transmission Line Fault Classification and Location Using Wavelet Entropy and Neural Network, Electrical Power Components and Systems, Vol.40, No.15, pp ,

6 [7] D.Chanda, N.K.Kishore, A.K.Sinha, Application of Wavelet Multiresolution Analysis for Classification of Faults on Transmission lines Electric Power Systems Research, 73, pp , [8] A. Abdollahi and S. Seyedtabaii, Transmission Line Fault Location Estimation by Fourier & Wavelet Transform Using ANN, The 4 th International Power Engineering and Optimization Conference (PEOCO), Shah Alam, Selangor, Malaysia, June [9] A.H. Osman and O.P.Malik, Transmission Line Distance Protection Based on Wavelet Transform, IEEE Transactions on Power Delivery, Vol. 19, No. 2, pp , April [10] R. Granizo, F. Álvarez, C.A. Platero, M. Redondo, Novel Protection Method for Ground Faults Detection in Cables Used in Combined Overhead- Cable Lines in Power Systems, IEEE Conference Publications.,13 July

7 967

8 968