Transmission Line Fault Diagnosis Using Chromatic Monitoring
Ziyad S. D. Almajali
In electrical power systems, the overhead transmission line (Figure 19.1 a) is utilised for facilitating power delivery of electrical energy to consumers despite the long distance separating it from its resources. Standard three phase parallel transmission lines increase the power transmission capability. A schematic diagram of parallel three phase power lines is shown in Figure 19.1b. Compared to the generation and distribution parts of the power system, high failure probability is recorded in the transmission part despite the high standard measures in the design and material selection.
In order to ensure a continuous and reliable supply of electric power, any faults occurring in such systems need to be detected urgently and the location and type of fault identified. Monitoring can only be done conveniently at both ends of the transmission line and the fault location and type determined from the source and receiving end signals.
Different faults may occur with large possibilities of different locations throughout the line. Monitoring is made more difficult because the voltage and current waveforms of each of the three phases may show complex variations upon fault occurrence. The variation depends on the severity of the fault and how close the fault is to the measurement points. Such complexity makes it difficult to directly interpret the waveform variation into useful information about the type of the fault which is causing it and the location.
The unknown and unpredictable fault parameters promote the use of artificial intelligence-based methods such as fuzzy logic (Ferrero et al., 1994; Kumar et al.. 1999) and neural networks (NNs) (Mohamed and Rao, 1995; Aggarwal et al., 2012) in addition to knowledge-based methods of wavelet transforms (Shaik and Pulipaka, 2015) and a combination of various methods (Reddy and Mohanta, 2007; Pothisarn and Ngaopitakkul, 2010; He et al., 2014).
Chromatic monitoring techniques have also been considered using computer-based simulation packages (Almajali et al., 2013, 2014). This has involved applying three chromatic processors to 60-Hz alternating current waveforms and extracting values of the resulting chromatic parameters to yield the required information.
FIGURE 19.1 Three phase electric power transmission lines, (a) Power system transmission lines, (b) Schematic diagram of parallel three phase power lines and a fault (F) [Ml = Source (S), М2 = Receiver (R)].
The available data for fault detection are in the form of two groups of sinusoidal waveforms, distinguished by the collecting ends of the transmission lines (sender or receiver) from w'hich they have been obtained. Each group of waveforms from each end contain another tw'o groups, three currents and three voltages from the three current phases (Ia, Ib, Ic; Figure 19.1b). Chromatic processing involves applying three chromatic processors (R. G, B) to these alternating current waveforms and extracting values of the resulting chromatic parameters (H, L, S) (Chapter 1).
In principle, all the available signals should be monitored, which would require a comprehensive and complex diagnostic system. However, a variation in the waveform characteristics due to a fault occurs not only in the faulty phase but also in the remaining phases, albeit reduced (Anderson, 1995). Therefore, the symmetrical components method is utilised as a pre-processing step to transform the three phase waveforms of both current and voltage into new different sets of waveforms. Thus, monitoring any one of the three symmetrical components avoids the need to monitor all three current components and reduces the calculation burden.
Figure 19.2a shows three typical alternating current (AC) waveforms corresponding to each of the three phase currents (Ia, Ib, Ic) at the receiving end of a transmission line. Phase Ib suffered a fault at a time of 33.5 ms.
The positive (Ial), negative (Ia2) and zero (Ia0) sequence components are calculated as follow's (Fortescue, 1918).
where a is a unit vector at an angle of 120°. a = 1Z1200 and a2 = 1Z2400.
FIGURE 19.2 Three phase current waveforms with fault, (a) Ia, Ib. Ic. (b) Positive symmetrical component (Ial).
An example of the positive sequence component (Ial) for Ia, Ib, Ic (Figure 19.2a) is shown in Figure 19.2b.
Chromatic Parameter Selection
The chromatic monitoring method is based on the continuous monitoring of vital signals from the power system (current and voltage waveforms) in the time domain. The process is applied to a series of time windows, each of a single cycle time duration. HLS chromatic parameters are calculated for each waveform from the outputs of the processors R, G, В (Chapter l). An example illustrating the RGB filtering implementations on a fully rectified post-fault cycle of a positive sequence waveform is shown in Figure 19.3.
FIGURE 19.3 An example of RGB processor application to a post-fault cycle of a rectified positive sequence component waveform.
A substantial number of chromatic parameters are produced from addressing the available waveforms of the sequence components by three chromatic processors (R. G and B) (Equations 19.1 through 19.3). The approach is then to select the parameters that are directly affected by the fault conditions for analysis and further processing. Secondary processing may then be used for connecting different sets of primary parameters for producing further useful parameters.
Chromatic processing has been used to monitor not only the occurrence of a fault but also the fault location and type of fault.
The objective of the fault locator is to provide rapid and accurate information about the fault location. This should not be affected by the type of fault, the fault resistance or any surrounding condition variations, even in the case of multiple fault conditions or any other possible disturbances. Fault location estimation can be made initially depending upon chromatic strength parameter (L) evaluation from the available R, G, В outputs for each of the rectified positive current symmetrical components and from both transmission line terminals.
To overcome any effects due to different fault resistances and so on, a secondary chromatic parameter (LRS) is introduced (Almajali et al., 2013). This parameter has been empirically determined
FIGURE 19.4 Variation of LRS over several cycles close to a fault, (a) At 75% of the line length, (b) at 50% of the line length, (c) at 35% of the line length with different fault resistances (Almajali, 2015).
as the relative strength difference between the source and receiver strengths (Lr- Ls) with respect to the sum of the source and receiver strengths (Lr + Ls). The parameter is given by Equation 19.4.
An example of the variation of LRS with time over several cycles close to a fault inception is given in Figure 19.4 for faults at different locations of the line.
The LRS value in Figure 19.4a reduces from unity at fault inception (cycle 1) to 0.75 at cycle 3, indicative of the fractional location of the fault halfway along the line length. LRS for faults at other line locations (0.5 of line length) is shown in Figure 19.4b, while Figure 19.4c illustrates how LRS transition is also independent of the fault electrical resistance by an example of a fault at 0.35 of line length but with different fault resistances (0.001, 100 and 200 Q). The method indicates a fault location to an accuracy of 1.9%, and this value is not affected by the fault electrical resistance or the position of the fault (Almajali et al„ 2013).
Figure 19.5 shows a flowchart for the complete process of the locator procedure, which involves required waveforms, in addition to the stages of data pre-processing, primary processing and secondary chromatic processing.
FIGURE 19.5 Locator flowchart (Almajali, 2015).
Identification of Fault Type
An indication of the fault type can be made chromatically to various levels of detail depending upon the chromatic parameters used. R.G,B outputs are available for each of the three current phases (Ia, Ib, Ic). The R. G, В outputs are processed to provide several H, L, S chromatic parameters. Figure
19.6 shows a flowchart which indicates the various combination of chromatic outputs and parameters which have been shown to provide various levels of fault types (Almajali et al„ 2014). Physically relevant fault types may be classified as belonging to one of three groups:
a. Faults between two of the three phases (AB, AC, BC)
b. Faults between each phase and ground (AG, BG, CG)
c. Faults between pairs of phases and ground (ABG, ACG, BCG)
where A, B, G refer to current phases Ia, Ib, Ic, respectively and G refers to the ground involvement. The flow' diagram indicates which chromatic parameters may be used to distinguish between the different faults.
An example of the fault type distinction w'hich can be achieved is the use of the H value of the negative symmetrical component Ia2 (Equation 19.2) of the alternating current (Figure 19.6).
FIGURE 19.6 Identification of the fault type algorithm flowchart (Almajali, 2015).
FIGURE 19.7 Chromatic analysis to indicate fault type: Ff from negative symmetrical component (Ia2) (Almajali, 2015).
Figure 19.7 shows the H parameter for Ia2 as a function of number of cycles for different types of faults at 50% of the transmission line length. The fault types cover all possible range of faults. The results show that various faults may be grouped according to different ranges of H but with some ambiguities; that is, the line-ground faults (ABG. ACG, BCG) are not distinguishable from the line-only faults (AB, AC, BC), and ambiguities can occur at the boundaries between each of the six ranges.
Limitations with the H parameter-based fault type identifier may be addressed through the use of the L parameter for the rectified zero symmetrical component Ia0 (Figure 19.6).
Figure 19.8 shows the L parameter for the rectified Ia0 as a function of number of cycles close to a fault of different types. This shows that the healthy condition together with the nine different faults form three clusters of L parameter levels. However, the non-ground and lines to ground faults are distinguishable by their different values of the chromatic L parameter.
If the H parameter information is combined with the L parameter information, ambiguities can be removed to provide a higher level of discrimination, as shown in the L of Ia0 versus H of Ia2 graph of Figure 19.9.
FIGURE 19.8 Chromatic analysis to indicate ground presence; L from rectified zero sequence component (Ia0) (fault resistance = 0.001 fi. fault location = 50%) (Almajali, 2015).
Summary and Overview
Chromatic techniques can be deployed for overhead transmission line fault diagnosis by identifying the fault type and location estimation. A combination of the signal strengths (Equation 19.4) of the rectified positive sequence components (Ial; Equation 19.1) forthe line at the source (Ls) and receiver (Lr) ends of a three phase power transmission line can identify the location of a fault along the line regardless of the electrical resistance of the fault (Figure 19.4).
The use of the chromatic parameter H of the negative sequence component (Ia2; Equation 19.2), along with the strength L for the rectified zero sequence component (Ia0) of an AC waveform enables various types of faults (single line to ground, line to line, line to ground) to be distinguished (Figure 19.9).
The chromatic approach provides a high level of transparency and traceability in monitoring electrical transmission line faults. Further development of the chromatic parameters available can lead to further fault discrimination capabilities.
FIGURE 19.9 Chromatic map of L(Ia0) versus H(Ial) with clusters of different faults (AG. BG. CG and АС/
ACG. BC/BCG. AB/ABG) plus normal condition (Norm) (Almajali, 2015).
Aggarwal R. K., Blond S. L., Beaumont R. Baber G., Kawano F.. and Miura S. (2012). High frequency fault location method for transmission lines based on artificial neural network and genetic algorithm using current signals only. IET Conference Publications, volume 2012, University of Bath.
Almajali Z. S. (2015). Fault Diagnosis for Transmission Lines Using Chromatic Processing. Ph.D. thesis University of Liverpool.
Almajali Z. S., Spencer J.W.. and Jones G.R. (2013). Fault Locator for a Parallel Transmission Line Using Chromatic Processing. The 8th Jordanian International Electrical and Electronics Engineering Conference, (JIEEEC 2013). Amman. Jordan.
Almajali Z. S., Spencer J.W., and Jones G.R. (2014). Asymmetrical Fault Classifier for a Parallel Transmission Line Using Chromatic Processing. The 7tli IET International Conference on Power Electronics, Machines and Drives (PEMD 2014). Manchester, UK.
Anderson P. M. (1995). Analysis of Faulted Power Systems [Electronic Book], IEEE Press Series on Power Engineering, Piscataway, NJ.
Ferrero A.. Sangiovanni S.. and Zappitelli E. (1994). A fuzzy-set approach to fault-type identification in digital relaying. Transmission and Distribution Conference, Proceedings of the 1994 IEEE Power Engineering Society, pp. 269-275.
Fortescue C. L. (1918). Method of symmetrical co-ordinates applied to the solution of polyphase networks. Proceedings of the American Institute of Electrical Engineers, 37(6): 629-716.
He Z., Lin S., Deng Y. J.. Li X. P. and Qian Q. Q. (2014). A rough membership neural network approach for fault classification in transmission lines. International Journal of Electrical Power and Energy Systems, 61:429-439.
Kumar P. Jamil M., Thomas M. S.. and Moinuddin (1999). Fuzzy approach to fault classification for transmission line protection. TENCON 99. Proceedings of the IEEE Region 10 Conference, volume 2, pp. 1046-1050.
Mohamed E. A. and Rao N. D. (1995). Artificial neural network based fault diagnostic system for electric power distribution feeders. Electric Power Systems Research, 35(1): 1-10.
Pothisarn C. and Ngaopitakkul A. (2010). Wavelet transform and fuzzy logic algorithm for fault location on double circuit transmission line. I6tli International Conference on Electrical Engineering (ICEE), Busan. Korea.
Reddy M. J. and Mohanta D. K. (2007). A wavelet-fuzzy combined approach for classification and location of transmission line faults. International Journal of Electrical Power and Energy Systems, 29: 669-678.
Shaik A. G. and Pulipaka R. R. V. (2015). A new wavelet based fault detection, classification and location in transmission lines. International Journal of Electrical Power and Energy Systems, 64: 35-40.