Harmonic Distortion in Transmission Networks due to Wind Farm Interconnection using IGBT Frequency Inverters Daphne Schwanz and Roberto Chouhy Leborgne Electric Power Systems Laboratory Federal University of Rio Grande do Sul Porto Alegre, Brazil daphne@ece.ufrgs.br and rcl@ece.ufrgs.br Abstract This paper presents an analysis of the harmonic distortion on the electric power system due to the connection of a large wind farm. The wind farm simulated use variable speed multipolar synchronous generators and their connection to the grid is made by an IGBT frequency inverter. These frequency inverters are modeled by using current sources. The simulation is performed in the time domain using the EMTP program. The current and voltage signals obtained at the wind farm point of common coupling (PCC) and the neighboring buses will be analyzed in order to calculate the total and individual harmonic distortion. The results will be showed by graphs and tables. The analysis of the voltage harmonic propagation into the bulk power system due to the connection of a new wind farm will show the possible need for harmonic distortion mitigation. I. INTRODUCTION The contribution of alternative sources of electricity generation, such as wind power, in the Brazilian energy matrix and worldwide has increased considerably in recent years. As a result, studies have been conducted to ascertain the impact that such energy sources can cause in their connection to the grid. The power quality is an important factor in these studies because it is related to any deviation in magnitude, waveform and frequency of the voltage or current, as the harmonic distortion. The harmonic distortion is caused by the operation of nonlinear loads and can cause problems for utility equipment such as overheating of transformers and failures of control equipment. The use of frequency inverters to connect wind farms to the grid can cause high harmonic distortion due to the switching of currents in these devices. Therefore, due to the huge increase of new wind farm projects to be connected in the power system of Rio Grande do Sul, it was carried out a case study on the harmonic propagation resulting from the connection of a wind farm located in Santana do Livramento. II. INERTERS AND FREQUENCY CONERTERS Frequency converters are built with semiconductor devices and intended for the control of frequency and voltage magnitude [1]. For variable speed wind turbines, the use of frequency converters is fundamental for the interconnection to the power system. Diodes and thyristors are the most frequently power electronic components used for rectifiers and inverters. Thyristor based frequency inverters are considered low-cost and low-loss devices. However, it consumes a large amount of reactive power and produces a considerable harmonic distortion [1]. These are some of the reasons for its limited use compared to GTO and IGBT inverters. Rectifiers using the latter technologies can be used for variable speed induction generators reactive power control [2]. Regarding frequency converters, there are different topologies that can be applied for wind farms, as back-to-back converters, multilevel converters and matrix converters. There are two types of switching strategies: the network switched converter and the auto-switched converter. In the first system, it is used a converter based on thyristors which consumes reactive power. Therefore, it is not able to control the reactive power. This type of system is used in high voltage and large power applications such as HDC systems [3]. In auto-switched systems, it is used PWM control methods and IGBT semiconductors. In this configuration it is possible to transfer active and reactive power in AC-DC (rectifier mode) or DC-AC (inverter mode). On the other hand, the PWM converter can produce high frequency harmonics and inter-harmonics. Fig. 1 shows the IGBT converter considered in this work. The authors also thank Capes and CNPq for financial support.
Figure 3. Example of a circuit using Thevenin and Norton Equivalents [4]. Figure 1. Frequency Converter using IGBT Technology. III. MODELING PROPOSAL According [4], [5] and [6] harmonics sources are nonlinear loads that can be modeled as current sources or voltage sources. The frequency converter used to connect the induction generator to the AC system is modeled in this study as a harmonic current source, as shown in Fig. 2. In order to use this model it is needed to have harmonic measurements or converter manufacturer data to calibrate the harmonic current sources. Another possible model is the Thevenin or Norton, which has a series or parallel impedance that moderates the response of the resonance of the circuit, as shown in Fig. 3. However, the determination of this impedance it is not straightforward, and often not very precise [4]. Thus, in this study it was used a current source model for the frequency converter, due to its simplicity without losing the reliability and accuracy of results. I. CASE STUDY The case study showed in this paper is based in the connection of the wind farm Cerro Chato, located in Santana do Livramento in Rio Grande do Sul. The connection of the wind farm to the grid was made by the bus LI2, as shown in Fig. 4. The single line diagram of the nearby power system is shown in Fig. 5. Fig. 6 shows the EMTP modeled wind generators and the step-up transformers. Figure 4. Wind Farm connection to the Grid. The study was carried out through time domain simulation using the program EMTP. This simulation approach is a very accurate method for harmonic analysis [7]. This wind generation facility is composed of three wind farms (Cerro Chato I, II and III). Each of them has a generation capacity of 30 MW (15 wind turbines of 2 MW), totaling 90 MW. The turbines are variable speed without gearbox, with a synchronous generator using a frequency converter for grid connection. Figure 2. Modeling of a Frequency Converter using a Current Source [4]. Figure 5. Single line diagram of the simulated network.
with two circuits. The parameters of this line are shown in Table IX. The voltage at the boundary buses were obtained by power flow simulation, as shown in Table X. The equivalent inductances of the buses MAC, URUG5, S.IC, P.MED and UTE URU are in Table XI. The loads at the buses LI2, ALE2 and BAG2 were modeled as constant impedances as shown in Table XII. The data of the substation power transformers are shown in Tables XIII to XI. These transformers were modeled using two different configurations. The bus ALE2 has just one 83MA transformer, as shown in Tables XIII and XI, and the buses LI2 and BAG2 have each two 69MA transformers as shown in Tables X and XI. The EMTP simulation was performed using a time window of 1s and a time step of 6μs. Figure 6. Equivalent of Wind Generators and Harmonics Sources The transmission system was modeled up to the third neighborhood bus from the point of common coupling (PAC) LI2. The rest of the grid was modeled by an equivalent circuit represented by a voltage source and the series shortcircuit impedance as shown in Fig. 5. The study was carried with the wind farm operating at its nominal power. According to [4], the frequency converters can be modeled by current sources. The amplitude of each one of the harmonic current sources was obtained by measurements carry out at the step-up 0.400/34.5k transformer. The RMS current of the fundamental and each harmonic frequency are shown in Table I. The data of the equivalent step-up transformer is shown in Table II. Using such procedure it is possible to simplify the modeling of the frequency converter without loss of reliability. The underground circuits and the transmission lines were modeled using frequency-dependent parameter models. For the transmission lines, it was used the J. Marti model, according to [8], which is widely used in the modeling of frequency-dependent overhead lines. The parameters used in this line model are shown in Table III and I. The underground cables connecting each generator were modeled by Wideband which is also a frequency dependent model [9]. The parameters used in this line model are shown in Table and I. The Substation Cerro Chato has three 34.5/230k power transformers. The parameters of these transformers are shown in Tables II and III. Furthermore, this substation is connected to the bus LI2 by a 24.83 km transmission line Harmonic Order TABLE I. HARMONIC SOURCES Current Harmonic (A) Order 1 2887 26 15 Current (A) 2 52 27 7.6 3 31 28 12 4 26 29 24 5 43 30 5.9 6 12 31 7.4 7 39 32 6.8 8 12 33 5.9 9 11 34 6.3 10 8.9 35 26 11 21 36 5.9 12 6.3 37 9.3 13 9.6 38 5.4 14 6.3 39 5.4 15 9.6 40 5.4 16 5.9 41 21 17 21 42 4.8 18 5.9 43 8.3 19 7.6 44 4.8 20 7.2 45 4.8 21 7.2 46 4.8 22 6.8 47 16 23 23 48 4.8 24 6.8 49 11 25 9.8 50 4.8
TABLE II. EQUIALENT STEP-UP TRANSFORMERS TABLE II. CERRO CHATO TRANSFORMERS Data Primary Secondary oltage 0,4 k 34,5 k Power of the Equivalent Transformer 10 MA 10 MA Connection Y Inductive Reactance 0.522 p.u. 0.522 p.u. Resistance 0.00522 p.u. 0.00522 p.u. TABLE III. Fase DC Resist. [ohm/km] GEOMETRIC DATA OF THE 230K TRANSMISSION LINES Ext. Diam. [cm] Horiz. Dist. Height of the Tower Height Fase 1 0.09 2.52 0 27.7 15 Fase 2 0.09 2.52 10.5 27.7 15 Fase 3 0.09 2.52 21 27.7 15 TABLE. TABLE I. Line LI 2 - ALE 2 LI 2 - BAG 2 ALE 2 MAÇ ALE 2 URU 5 ALE 2 S.IC ALE 2 UTE URU BAG 2 P. MÉD. Cable Type LENGTH OF THE 230K TRANSMISSION LINES Length 123 km 171 km 80 km 146 km 115 km 143 km 52.3 km GEOMETRIC DATA OF THE UNDERGROUND CABLES OF THE WIND FARM Single Core Number of Cables 3 ertical Distance Horizontal Distance between Cables External Radius of the Cable Internal Radius of the Conductor External Radius of the Conductor Internal Radius of Aluminum Coating External Radius of Aluminum Coating TABLE I. 0.9 m 0 0.2 0.4 m 0.02 m 0 m 6.42E-3 m 0.01522 m 0.018 m LENGTH OF THE UNDERGROUND CABLES Line Circuit Length Cerro Chato I Cerro Chato II Cerro Chato III Circuit 1 Circuit 1 Circuit 1 3.14 km 2.58 km 1.19 km 4.18 km 2.89 km 1.70 km 1.42 km 1.44 km 2.05 km TABLE III. Data Primary Secondary oltage 34.5 k 230 k Power 35 MA 35 MA Connection Y Y IMPEDANCE OF THE SUBSTATION TRANSFORMERS Transformer Impedance Primary Secondary CCH 1 Cerro Chato Substation CCH 2 Cerro Chato Substation CCH 3 Cerro Chato Substation TABLE IX. Fase DC Resist. [ohm/km] Reactance 0.2627 pu 0.2627 pu Resistance 0.01018 pu 0.01018 pu Reactance 0.2757 pu 0.2757 pu Resistance 0.01010 pu 0.01010 pu Reactance 0.2789 pu 0.2789 pu Resistance 0.01031 pu 0.01031 pu GEOMETRIC DATA OF THE 230K TRANSMISSION LINE CERRO CHATO - LI2 Ext. Diam. [cm] Horiz. Dist. Height of the Tower Height Fase 1 0.0917 2.5146-3.9 32.2 20.2 Fase 1 0.0917 2.5146 3.9 32.2 20.2 Fase 2 0.0917 2.5146-39 32.2 15.2 Fase 2 0.0917 2.5146 3.9 32.2 15.2 Fase 3 0.0917 2.5146-3.9 322 10.2 Fase 4 0.0917 2.5146 3.9 32.2 102 Fase 0 4.1889 0.914-3 37.5 30.9 Fase 0 0.329 1.5 3 37.5 30.9 TABLE X. OLTAGES AND ANGLES IN THE ANALYZED BUSES oltage (k) Angle (degrees) MAC 230-14 URUG 5 230-15 S.IC 230-16 P. MED 230-8.2 UTE URU 230-15 TABLE XI. SHORT-CIRCUIT EQUIALENT INDUCTANCES Equivalent Inductance (mh) MAC 70.6 mh URUG 5 55.2 mh S.IC 99.3 mh P. MED 30.9 mh UTE URU 49.6 mh TABLE XII. LOADS CONNECTED TO THE BUSES Resistance (Ω) Inductance (mh) LI2 433 3608 ALE2 381 ----- BAG2 397 1052
TABLE XIII. TRASFORMER AT ALE2 Data Primary Secondary Tertiary oltage 230 k 69 k 13.8 k Power 83 MA 83 MA 83 MA Connection Y Y TABLE XI. Resistance Inductive Reactance TABLE X. IMPEDANCE OF THE TRASFORMER AT ALE2 R 12 R 13 R 23 4.337E-3 pu 7.2289E-3 pu 6.5070E-3 pu X 12 X 13 X 23 0.1577 pu 0.2398 pu 0.0639 pu TRASFORMERS AT LI2 AND BAG2 Data Primary Secondary tertiary oltage 230 k 69 k 13.8 k Power 69 MA 69 MA 69 MA Connection Y Y for IHD and THD are shown in Table XII and XIII, respectively. According to the results shown in Table XII, it was observed that the harmonic propagation was not the same for each frequency. Some frequencies were dumped and other amplified as a result of transmission network impedances. For same frequencies resonance between transmission lines series inductance and shunt capacitance may happened. For most frequencies there was a reduction in the value of the individual harmonic distortion as the observation point distanced from the harmonic source, the distance of the nearby buses from the Wind Farm are shown in Table XIX. The total harmonic distortion at the three buses analyzed is in accordance with the recommended limits [10], as shown in Table XX. TABLE XI. Resistance Inductive Reactance IMPEDANCES OF TRASFORMERS AT LI2 AND BAG2 R 12 R 13 R 23 6.4E-3 pu 0.014 pu 0.0118 pu X 12 X 13 X 23 0.2674 pu 0.458 pu 0.1721 pu TABLE XII. INDIIDUAL OLTAGE HARMONIC DISTORTION Frequency (Hz) 300 IHD(%) LI2 0.22 BAG2 0.09 ALE2 0.09. HARMONIC DISTORTION AND PROPAGATION The harmonic propagation analysis was performed by comparing the results obtained from the individual and total harmonic distortion of the voltage at LI2, ALE2 and BAG2, using (1) and (2), respectively. IHD = h 100, THD = H 1 h= 2 1 2 h 100 Where h is the h-harmonic voltage and 1 the fundamental frequency voltage. The harmonic propagation analysis is of paramount importance, because it verifies how far the harmonic frequencies propagate. Considering that only the wind farm is injecting harmonic currents in the electrical system, it was possible to determine the rate of propagation of harmonic according to the voltage level and the distance from the source harmonic. I. RESULTS It was calculated the individual harmonic distortion (IHD) and total harmonic distortion (THD) using (1) and (2). The total harmonic distortion was calculated considering up to the 50th harmonic frequency. Only the most representative harmonic frequencies are shown in Table XII. The results LI2 0.07 420 BAG2 0.04 ALE2 0.08 LI2 0.02 660 BAG2 0.07 ALE2 0.005 LI2 0.0004 1020 BAG2 0.0004 ALE2 0.0004 LI2 0.04 1740 BAG2 0.04 ALE2 0.0008 TABLE XIII. TOTAL OLTAGE HARMONIC DISTORTION THD(%) LI2 0.42 BAG2 0.46 ALE2 0.29 TABLE XIX. DISTANCE FROM THE HARMONIC SOURCE (WIND FARM) Distance (km) LI2 24.8 BAG2 195.8 ALE2 147.8
TABLE XX. oltage HARMONIC DISTORTION LIMITS Total Harmonic Limit (%) 69 k 5 69,001 k 161 k 2.5 161,001 k 1.5 II. CONCLUSION Using the proposed methodology of simulation, it was possible to determine the harmonic propagation caused by the connection of a large wind farm through IGBT frequency converters. It was observed that the total harmonic distortion exceeded the maximum recommended values in all the analyzed buses. The propagation of the harmonics, assessed by the individual and total distortion indices, showed a heterogeneous behavior. For some frequencies the distortion was dumped while for other it was amplified. From the results it is possible to conclude that harmonic distortion should be a great concern for power system operators when large wind farms apply for connection permits. ACKNOWLEDGMENT The authors thank engineers Miguel Pires De Carli, Renato Ferraz Gonçalves, Yuri Solis Stypulkowski, Leonardo Ulises Iurinic, Martin Cruz Rodriguez Paz, Daniel da Silva Gazzana e André Bernardes Michel by the data provided and knowledge shared. REFERENCES [1] Arrilaga, J., Watson N. R., Power System Harmonics, John Wiley & Sons, 2nd edition, New Zeland, 2003. [2] Ackermann, T., Wind Power in Power Systems, John Wiley & Sons, 2 nd edition, England, 2005. [3] Chen, Z., Guerrero, J. M., Blaabjerg, F., A Review of the State of the Art of Power Electronics for Wind Turbines, IEEE Transactions on Power Electronics, ol. 24, N 8, August, 2009. [4] Dugan, R. C., McGranaghan, M. F., Santoso, S. Beaty., H. W., Electrical Power Systems Quality, McGraw-Hill, 2 nd edition, 2004. [5] Pomilio, J. A., Deckmann, S. M., Characterization and Compensation of Harmonics and Reactive Power of Residential and Comercial Loads, IEEE Transactions on Power Delivery, ol. 22, N 2, April, 2007. [6] Oliveira, L.C.O., Melo, G. A., Souza, B. D., Canesin, C. A., Bonatto, J. B., Belchior, F. N., Oliveira, M., Mertens Jr., E. A., Harmonic Propagation Analysis in Electric Energy Distribution Systems, 11 th International Con Electrical Power Quality and Utilisation (EPQU), 2011. [7] Badrzadeh, B., Gupta, M.,Singh N., Petersson, A., Max, L., Hogdahl, M., Power System Analysis in Wind Power Plants Part I: Study Methodology and Techniques, IEEE Industry Applications Society Annual Meeting (IAS), 2012. [8] Marti, J., Accurate Modelling of Frequency-Dependent Transmission Lines in Electromagnetic Transient Simulation, IEEE Transactions on Power Apparatus and Systems, ol. PAS-101, Nº 1, January 1982. [9] Morched, A., Gustavsen, B., Tartibi, M., A Universal Model for Accurate Calculation of Electromagnetic Transients on Overhead Lines and Underground Cable, IEEE Transactions on Power Delivery, ol. 14, N 3, July 1999. [10] International Electrotechnical Commission. IEC/TR 61000-3-6: Electromagnetic compatibility (EMC) - Part 38: Limits - Assessment of emission limits for the connection of distorting installations to M, H and EH power systems. International Electrotechnical Commission, Genebra: 2008. 58 p.