HARMONIC DISTURBANCE COMPENSATING AND MONITORING IN ELECTRIC TRACTION SYSTEM

HARMONIC DISTURBANCE COMPENSATING AND MONITORING IN ELECTRIC TRACTION SYSTEM A. J. Ghanizadeh, S. H. Hosseinian, G. B. Gharehpetian Electrical Engineering Department, Amirkabir University of Technology, Tehran, Iran (ghanizadeh/hosseinian/grptian)@aut.ac.ir ABSTRACT This paper presents methods to compensate and monitor the disturbance generated by electric traction units, considering their position in different times, in a railway system. For this purpose, first of all, experimental data of Italy traction system is used and its components are modeled. Then, using proposed algorithm a suitable location for a passive filter is determined. Subsequently, considering Speed of traction units, the location of each traction unit is predicted and with proposed THD index, created disturbance in each time is determined. Finally, in the span of 1 minute, proposed index is monitored. It is worth mentioning that the Matlab / Simulink software is used to simulate proposed model. KEYWORDS Total Harmonic Distortion, Power quality, Electric traction system, Monitoring, Harmonic filtering 1. INTRODUCTION With increasing the number of non-linear loads, the main requirement of any power system is to supply electricity with a determined power quality and reliability to minimize possible cost. As well known, traction systems can cause distortion and harmonic in shape of voltage and current waves and these harmonics can lead to unacceptable effects like energy loss, communication interference and the like. As a result, attention has to be paid to improve their operations. In literature, there are attentions in this serious problem. In [1-3], effect of electric traction systems is analyzed on power quality. In references [4, 5], interaction between AC and DC electric railway systems, which cause turbulence and reduce the quality of the network, has been studied. In addition, reference [4] investigates the use of AC/DC converters in electric railway system that causes voltage and current distortions. In [6a], a DC electric traction system has been modelled and Static Var Compensator (SVC) and a filter was added to the distribution network to improve the power quality problems. The interconnection between the new European railway lines (25 kv/50 Hz) and the existing DC 3kV Italian railway lines has been presented in [7a]. Analysis of the filters has been done in order to mitigate the effects between the two systems. A harmonic study for a simplified model of line in a traction system without considering earth effect is done in [8]. Subsequently, reference [9] proposed a better model for transmission line considering earth return path with time domain modeling. So far, no study on DC traction system is done that be able to analyze Total Harmonic Distortion (THD) index with considering speed and position of traction units and all components (filters, feed lines and motors) are modeled completely. In this paper firstly, using Matlab/Simulink software, all components of traction system are modeled and then using proposed algorithm a suitable location for a passive filter is determined. Subsequently, considering speed of traction units, the location of each traction unit is predicted and THD index is monitored in each bus of traction system. It is worth mentioning that THD 1

index is monitored over the time of 1 minute in different states. The First state is when there is no filter in traction system. The second state is when passive filter, determined by proposed algorithm, is located on proper location. Finally, the third state is when all traction units have their suitable filter. This paper is organized as follows: in the next section, different components of Italy traction system are modeled using Matlab/Simulink software and proposed algorithm to determine suitable location of passive filter is defined. Finally, in section 3, simulation result and analysis is shown. 2. DESCRIPTION OF THE TRACTION SYSTEM The electrical power system in study is a part of electric traction system in Italy [4]. This system is shown in Fig. 1. This system is fed by a 380 kv, 1000 MVA bus. Also, this system consist of a AC traction unit, 25 kv, 50 Hz, and four DC traction units, 3 kv and with 6 pulse AC/DC converters 132 kv, 5.4 MW. Moreover, total active power absorbed by AC traction station and each DC traction stations are equal to 54 MW and 4.1 MW respectively. Figure 1. Three-phase transmission system along with AC and DC electric railway systems Subsystem of AC and DC traction units are shown in Fig. 2. Moreover data of lines, transformers and linear loads are listed in table 1, 2 and 3. 2

Figure 2. (a) DC traction system subsystem, (b) AC traction system subsystem Lines Line V n r x c L Type (kv) (Ω/km) (Ω/km) (nf/km) (km) 1-2 1-3 1-7 3-4 4-5 AT-2 Cable Overhead Cable Overhead Overhead Cable 132 132 20 132 132 132 0.0291 0.172 0.04 0.172 0.172 0.0291 0.122 0.424 0.046 0.424 0.424 0.122 220 8.5 700 8.5 8.5 220 3.85 24 3 17 18 8.16 Table 1. Lines data Trans. S n V 1/V 2 r x (MVA) (kv/kv) (%) (%) AT 250 380/132 0.2 13 T1 25 132/20 0.8 14 T2 25 132/20 0.8 14 T3 25 132/20 0.8 14 Table 2. Transformer data Bus Vn Pn Qn (kv) (MW) (MVar) 6 20 8 6 7 20 20 21.8 Table 3. Linear loads data 2.1. DC line model Considering the movement of train between two stations, amount of electric parameters of DC line are changing perpetually. As a result, for accurate analysis of power quality index in traction system, it is needed to model these changes completely. According Fig. 3, train in each time, is located in special position with special speed. 3

Figure 3. Spatial status between the two train stations To model position and speed of railway in power quality calculation, first of all, parameters per length of DC line should be determined. Table 4 listed these parameters. l= Ɩ /L r = R/L c = C/L 1.37 µh/m 0.5 µω/m 20 pf/m Table 4. Electrical parameters of line Now with these parameters and speed characteristic of train that shown in shown in Fig. 4, model of DC line can be obtained. As Fig. 4, shown, it is assumed that railway start moving from first station and with accelerating movement arrive to maximum speed after t1 seconds. Then, continue with the same speed until t2 that railway reduces its speed to stop at the second station at tmax. To model the position of train and the effect of DC line in simulation, every moment, the train speed and distance traveled by it computes. After that, considering amount of length that is determined by speed of train and Table 4, electric parameters of DC line are determined. Since the distance between stations, about 1.5 km is assumed, the appropriate model for the DC line, pi model is considered. 2.2. Motor model Figure 4. Speed characteristics of the train sets with time variations There are different kinds of DC electric motors that feed by DC current. Series motors have high initial torque, hence suitable for electric railways. In addition, in equal load, series motors consume fewer current in comparing with other electric motors. In traction systems, both series and independent excitation motors are used. In [4], railway motors are modeled by resistor. It is clear that this model is so simple and it is not reliable. In this paper, a perfect model is used. As mentioned, a series motor is used to model 4

railway motor. In addition, considering speed characteristic of railway, speed of motor in each time is given as input to the model of motor. 2.3. Harmonic filtering Using power electronic converter in electric transportation systems such as electric trains causes a lot of harmonic levels inject to distribution networks. Harmonic filters are used to limit the harmonic currents flowing into the upstream network. The filter could be located on the loco itself or at the substation. Because of different type of trains running simultaneously on the same traction section move in a section, the best place can be offered to install them, are traction posts [10]. Passive, active and their mixture (hybrid) are three kinds of filters. Passive filters as classic methods for power quality improvement of distribution systems consist of series LC tuned for removing a specific harmonic or blocking a bandwidth of severe harmonics of nonlinear loads current. These filters have low impedances for the tuned frequencies such as 5th and 7th. Low cost is a great benefit of these filters. To reduce the overall amount of distortion, after calculating the amount of THD index at every bus of system, the proposed algorithm, choose two higher harmonic to set parameters of filter. With considering the location of filter, three states are discussed in this paper. Firstly, using proposed algorithm, the proper location for filter in one station is determined and THD index is monitored in each bus. In Second state, without any filter in system, THD index in all buses are monitored. Finally, with locating filter in each railway, THD index in third state is monitor in each bus. 2.4. Description of the Matlab model The power system block set of Matlab/Simulink is design tools that can be build the simulation models for electrical power system. Part of Simulink model of whole system which consists of DC and AC electric substation (ESS) is proposed in Fig. 5. The model that has been used in the train simulation that consists of inverter and DC line and DC machine is shown in Fig. 6. 3. SIMULATION RESULTS Harmonic has undesirable effects in the network, such as reduction in transformers efficiency and more losses in motors; therefore, identification of the disturbance and determination amount of them and finally its reduction in the electric train system is important. In this section several simulations on the system in figure.1, has been done. Let the maximum speed of train be 100 km/h and the radius of the wheels of train is 0.406m. t1, t2 and t3 respectively 5, 55 and 60 seconds have been considered. Sampling time is 5*10-5. In the first simulation, motor of train and DC line is modeled with resistance and assumed that no filter is installed. 5

Figure 5. Part of traction system simulink model The value of THD on each bus has been calculated. In order to perform the harmonic analysis of voltage and current signals present in the traction system, the proposed algorithm, shows graphically the harmonic spectrum of the analyzed signal. For example, the voltage of bus no. 5 (phase A), and the amount of THD and harmonic spectrum is given in Fig. 7. As shown in this figure, the value of 5th and 7th harmonics are10.696% and 8.297% respectively. The value of THD% in other buses of system is given in Table 5. Figure 6. The whole simulink model of DC-ESS 6

1 V5 (phase A) 0.5 0-0.5 Mag (% of 60 Hz component) -1 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 time (s) 10 Fundamental (60 Hz)= 0.961803 THD= 13.6133 8 6 4 2 0 0 100 200 300 400 500 600 700 Frequency (Hz) Figure 7. Voltage of bus 5 and its harmonic spectrum phase bus1 bus2 bus3 bus4 bus5 bus6 bus7 A 10.024 9.924 12.044 13.084 13.613 10.479 10.491 B 11.685 11.543 14.365 15.665 16.323 8.863 8.871 C 13.388 13.258 15.731 16.850 17.406 12.163 12.174 Table 5. The amount of THD% for other Buses In the next step, to reduce the amount of THD index, the proposed algorithm chooses two higher harmonic to set parameters of filter and then placed it at different buses. Results are shown in Table 6. It is noted that the installation of filter reduces the amount of THD. If the average amount of total harmonic distortion at all buses, for each test as a criterion for measuring the different scenario to be considered, best achieved when the filter is intended to be installed in the bus no. 3. That improves the value of THD from 12.759% to 3.65%. In the next simulations, a perfect model is used to simulate the DC line and motor of train. To see their effect and monitor the value of THD% at any moment in time, Assumed that train will move toward from one station to another station. For example, the amount of THD for bus no.1 (phase B), at different moments of train moving, is shown in Fig. 8. Position of passive filter bus phase Without filter bus1 bus2 bus3 bus4 bus5 bus6 bus7 1 2 3 A 10.0235 1.64037 1.62761 2.45633 3.92694 4.57236 9.41162 9.59333 B 11.685 3.01672 3.03987 3.49987 3.81001 4.15739 9.68616 9.92667 C 13.3882 4.10133 4.07718 4.59243 5.88083 5.73307 11.3269 11.6200 A 9.9236 1.61876 1.60169 2.44448 3.94148 4.56619 9.3311 9.51086 B 11.5434 2.96903 3.00251 3.51618 3.8289 4.14885 9.57276 9.80969 C 13.2575 4.0327 4.04721 4.59865 5.91932 5.72974 11.2279 11.5177 A 12.0438 3.76292 3.95992 2.6075 3.39185 4.42345 11.7456 11.8828 B 14.3654 5.10691 5.30555 3.34679 3.55694 4.08937 12.6588 12.8921 C 15.7311 5.61073 5.69912 4.67579 5.24526 5.64176 13.9325 14.2012 7

A 13.0836 5.18738 5.41696 3.06023 3.30901 4.18818 12.9098 13.0258 4 B 15.6649 6.59954 6.83678 3.97497 3.76406 3.91534 14.1113 14.3364 C 16.8504 6.79105 6.9038 5.1680 5.42461 5.64513 15.1844 15.4394 A 13.6133 5.93366 6.17495 3.53583 3.44305 4.20007 13.4974 13.6021 5 B 16.3225 7.40997 7.66147 4.5263 3.90406 4.01243 14.8521 15.0709 C 17.4057 7.45839 7.57841 5.58978 5.56641 5.91137 15.8098 16.0565 A 10.4787 3.2043 3.1544 3.48213 4.47295 4.79036 2.54488 2.6123 6 B 8.86261 1.29654 1.32918 1.91481 2.52157 3.39341 3.75009 3.40301 C 12.1628 4.21491 4.19796 4.12115 4.59194 4.66916 4.98671 4.6568 A 10.4907 3.21498 3.16414 3.4965 4.48288 4.79907 2.55052 2.43059 7 B 8.87125 1.2986 1.33186 1.91879 2.52799 3.40076 3.75392 3.56571 C 12.1741 4.22785 4.21067 4.13379 4.60088 4.67655 4.99271 4.81052 Avg 12.75915 4.22365 4.301011 3.65049 4.195759 4.603048 9.896999 9.998304 Table 6.The value of THD% at different buses of system, for different scenario of passive filter position 40 Instantaneous Voltage THD(%) for Bus1 35 30 THD% 25 20 15 10 0 10 20 30 40 50 60 Time(s) Figure 8. THD index at different moments of moving train in the span of 1 minute (no filter) Considering that the number of simulated cases is high, only some of the state is expressed. Fig. 9 represents the results for the prior case with the difference that the filter is installed in system. The results for other buses in system for the recent case are shown in Fig. 10. Variable and nonlinear characteristics of the electric railway system always cause harmonic distortion, especially in the beginning to move and stop. 8 Instantaneous Voltage THD(%) for Bus1 7 6 THD(%) 5 4 3 2 0 10 20 30 40 50 60 Time(s) Figure 9. THD index at different moments of moving train in the span of 1 minute (with filter) 8

Figure 10. Instantaneous THD at different Buses in the span of 1 minute (with filter) Hence one basic and inexpensive way to improve the power quality is using passive filter. If the parameter of this filter is selected as optimal, then this approach has required efficiency. Electric railway system is complex, as it includes electronic converter along with systems like track, electrical motor, control and mechanical parts. Future work will be devoted to a more detailed modeling of the train, simultaneous movement of two trains in a section and using of active filter for improving power quality. 4. CONCLUSION Harmonic has undesirable effects in the network, such as reduction in transformers efficiency and more losses in motors; therefore, identification of the disturbance and determination amount of them and finally its reduction in the electric train system is important. For evaluation of system performance considering the power quality, it is necessary to select appropriate indicators. In this paper, the THD harmonic distortion index, which is famous for the power quality studies, was used. Due to changes in train speed and location at different moments, the value of the index is change. In this paper, considering speed of traction units, the location of each traction unit is 9

predicted and THD index is monitored in each bus of traction system at each moment of time. Subsequently, proposed algorithm trying to improve the THD using the appropriate filter set and suitable location of passive filter is defined. REFERENCES [1] A. Capasso: The Power Quality Concern in Railway Electrification Studies., Proceedings of 8th IEEE PES International Conference on Harmonics and Quality of Power, Athens (Greece), Vol. 2, 1998, pp. 647-652. [2] R.E. Morrison: Power Quality Issues on AC Traction systems., Proceedings of 9th IEEE PES International conference on Harmonics and Quality of Power, Orlando (USA), Vol. 2, Oct. 2000, pp. 709-714. [3] M. Brenna, F. Foiadelli, and D. Zaninelli: Electromagnetic model of high speed railway lines for power quality studies., IEEE Transactions on Power System, Vol. 25, No. 3, 2010, pp. 1301 1308. [4] L. Battistelli, P. Caramia, G. Carpinelli, D Proto: Power Quality Disturbance Due to Interaction between AC and DC Traction System., Proceedings of the IEE International Conference on Power Electronics, Machine and Drives PEMD, Edinburge, April, 2004, pp.492-497. [5] L. Battistelli, P. Caramia, G. Carpinelli, D Lauria, D. Proto: A Power Quality Compensation Device for Interacting AC-DC Railway Systems., IEEE Power Tech Conference, St. Petersburg, 27-30 Jun 2005, pp. 1-6. [6] J. Martinez, G. Ramos: Reactive Power and Harmonic Distortion Control in Electric Traction Systems., 2010 IEEE/PES Transmission and Distribution Conference and Exposition: Latin America, Sao Paulo, Brazil, 8-10 Nov. 2010, pp. 190-195. [7] M. Brenna, F. Foiadelli: Analysis of the Filters Installed in the Interconnection Points Between Different Railway Supply Systems., IEEE Transactions on Smart Grid, Vol. 3, No. 1, March 2012, pp. 551-558. [8] J. Holtz and H.J. Klein: The propagation of harmonic currents generated by inverter-fed locomotives in the distributed overhead supply system., IEEE Transactions on Power Electronics, Vol. 4, No. 2, April 1989, pp. 168-174. [9] P.C. Coles, M. Fracchia, R.J. Hill, P. Pozzobon and G. Sciutto: Modeling and simulation of supply current interference in traction systems arising from multi-level converters in high-power locomotive drives., Proc. of 5th European Conference on Power Electronics and Applications, Brighton UK, Vol. 7, 13-16 September 1993, pp. 24-28. [10] P. Kiss and A. Dán: The Application of the Double Domain Simulation by Different Harmonic Filtering Methods of 25 kv Electric Traction Systems., Proceedings of 13th IEEE International conference on Harmonics and Quality of Power, Vol., 2008, pp. 1-6. Authors information Ahmad Javid Ghanizadeh was born in Herat, Afghanistan, in 1981. He received his B.Sc. and M.Sc. in electrical engineering from Sahand University of Technology, Tabriz, Iran and Ferdowsi University of Mashhad, Mashhad, Iran, in 2004 and 2009, respectively. He is currently pursuing the Ph.D. degree in the electrical engineering department at the Amirkabir University of Technology (AUT), Tehran, Iran. His research interests include Power Quality, power system optimization and operation and transformers transients. Seyed Hossein Hosseinian was born in 1961 in Iran. He received both the B.Sc. and M.Sc. degrees from the Electrical Eng. Dept. of Amirkabir University of technology, Iran, in 1985, and 1988, respectively, and PhD degree in Electrical Engineering Dept, university of Newcastle England, 1995. At the present, he is the associate Professor of Electrical engineering Department at the Amirkabir University of Technology (AUT). His especial fields of interest include, Power Quality, Restructuring and Deregulation in power systems. 10

Gevork. B. Gharehpetian received his BS, MS and Ph.D. degrees in electrical engineering in 1987, 1989 and 1996 from Tabriz University, Tabriz, Iran and Amirkabir University of Technology (AUT), Tehran, Iran and Tehran University, Tehran, Iran, respectively, graduating all with First Class Honors. As a Ph.D. student, he has received scholarship from DAAD (German Academic Exchange Service) from 1993 to 1996 and he was with High Voltage Institute of RWTH Aachen, Aachen, Germany. He has been holding the Assistant Professor position at AUT from 1997 to 2003, the position of Associate Professor from 2004 to 2007 and has been Professor since 2007. The power engineering group of AUT has been selected as a Center of Excellence on Power Systems in Iran since 2001. He is a member of this center. He was selected by the ministry of higher education as the distinguished professor of Iran and by IAEEE (Iranian Association of Electrical and Electronics Engineers) as the distinguished researcher of Iran and was awarded the National Prize in 2008 and 2010, respectively. Prof. Gharehpetian is a senior and distinguished member of IEEE and IAEEE, respectively, and a member of the central board of IAEEE. Since 2004, he is the Editor-in-Chief of the Journal of IAEEE. He is the author of more than 500 journal and conference papers. His teaching and research interest include power system and transformers transients and power electronics applications in power systems. 11