MODELING THE ELECTROMAGNETIC POLLUTION OF THE ELECTRIC ARC FURNACES

MODELING THE ELECTROMAGNETIC POLLUTION OF THE ELECTRIC ARC FURNACES MANUELA PĂNOIU 1, CAIUS PĂNOIU 2, IOAN ŞORA 3 Key words: Power quality, Electromagnetic pollution, Harmonics, Electric arc furnace (EAF), Electric arc modeling. The electric arc furnace is a three-phase load representing one of the most important generator of harmonic currents, reactive power and unbalanced conditions in electrical power systems. This paper aims to achieve a study concerning the power quality of the energy at 3-phase electric arc furnaces. There are analyzed the main qualitative indicators of the electric power: power factor, reactive factor, distorting factor and the total harmonic distortion for current and voltage to a 100 t EAF from an industrial platform. There are also study the modeling of a three phase electric arc furnace installation. Thus, for modeling his behavior, it was used a model which parameters like of a real electric arc. The simulations results are comparing with the measurements. 1. INTRODUCTION Electric arc furnace is a massive generator of disturbing in electrical power system. The three types of disturbing the electrical power system are: generation of the three phased harmonic currents, of an important reactive power and unbalanced high power three-phase load. In order to improve the electric power s qualitative indicators, it is necessary to use some solutions or specialized equipment to filter the harmonic currents, to improve the power factor and to balance the 3-phase electric charge [1 4]. The effect of these installations was analyzed using simulation program PSCAD/EMTDC [13]. PSCAD (Power System Computer Aided Design) is a graphical user interface capable of supporting a variety of power system simulation programs. This release supports EMTDC (Electro- Magnetic Transients in DC Systems). 1 Polytechnic University of Timişoara, Engineering Faculty of Hunedoara, Revoluţiei 4, 331128; E-mail: m.panoiu@fih.upt.ro 2 Polytechnic University of Timişoara, Engineering Faculty of Hunedoara, Revoluţiei 4, 331128; E-mail: c.panoiu@fih.upt.ro 3 Polytechnic University of Timişoara, Faculty of Electrical Engineering, Timişoara, B-dul V. Pârvan No 2, cod 300223, E-mail: ioan.sora@et.upt.ro Rev. Roum. Sci. Techn. Électrotechn. et Énerg., 54, 2, p. 165 174, Bucarest, 2009

166 Manuela Pănoiu, Caius Pănoiu, Ioan Şora 2 2. THE POWER QUALITY ANALYSIS The criteria of quantitative analysis of power quality are [1 4]: The power factor 2 2 2 = P S = P P + Q D (1) K P + The reactive factor and respectively the deforming factor K Q = Q P, (2) 2 2 = D P Q, (3) K D + where P is active power, S is apparent power, Q is reactive power and D is distorting power [2, 3]. The total harmonic distorsion for current and voltage (%) 2 40 40 Ik U k THDI = 100, THDU = 100, k = 2 I1 k = 2 U (4) 1 where U 1, I 1 are the rms values for voltage and current on the fundamental and U k, I k the effective values for current and voltage on the k-th harmonic order. The negative nonsimetry factors for voltage and current (%): + k U = U U 100, k I = I I 100. (5) The zero nonsimetry factors for voltage and current (%): 0 0 + 0 0 + k U = U U 100, k I = I I 100, (6) where U and I are the inverse sequences components, 0 sequences components and U and and currents. + 2 + U and + I are the direct 0 I are the homopolare components of voltages 3. RESULTS OBTAINED BY MEASUREMENTS IN THE ELECTRIC INSTALLATION OF A 100 t EAF The measurements were made at a 3-phase power supply installation of a 3-phase EAF of 100 t. It s been used a computer system with an ADA3100 [7] data acquisition board. The data acquisition on the 6 channels was made as was present in [14].

3 Modeling the electromagnetic pollution of the electric arc furnaces 167 3.1. THE MEASUREMENTS ON THE LOW VOLTAGE LINE As regards to the wave forms of the currents and voltages on the low voltage supply line, presented in Fig. 1, in the melting phase is found a strong distorsion of these. Also, one can notice that the load is strongly unbalanced. In the phase of the electric arc s stable burning, is found that the distortions that appear in the current s and voltage s wave forms are more reduced, as was show in [14]. Fig. 1 The variation for voltage and current during the melting stage. The spectral characteristics of the current and voltage were achieved by processing the data acquired by using FFT (Fast Fourier Transform) and shown in [14]. It was also calculate the powers and power factors, the total harmonic distortions, as well as the active and reactive powers on each harmonic, based on the equations known from electrotechnics [2] and the equations (1) (6). It s been obtained the variation form of all the measured values presented on the heat s entire duration and shown in [14]. From the total harmonic distortion was found values within 5 25 % for voltage, respectively 5 35 %, for the current. 3.2. THE MEASUREMENTS ON THE MEDIUM VOLTAGE LINE The wave forms of the currents and voltages from the medium voltage supply line are shown in [14]. Is found both the presence of a strong distortion of these and an unbalance due to the inequality of the amplitudes from the 3 phases. The power factor s value is within 0.5 0.9, the higher value being reached during the electric arc s stable burning and the lower value during the melting phase. In Fig. 2 is presented the variation form of the total harmonic distortion of the current, respectively the voltage, on the entire heat s duration. The distortion s value is

168 Manuela Pănoiu, Caius Pănoiu, Ioan Şora 4 higher in case of the current s wave than the case of the voltage s one. Were obtained values for the total harmonic distortion within 8 12 % for voltage, respectively 10 40 % for current. Comparing these values with the standards [8 10] where value of the permitted THD on the medium voltage lines is of 8 %, results also that the analyzed installation is not matched in the international standards. Fig. 2 The variation of the total harmonic distortion of the current and voltage, on the entire heat s duration on the medium voltage line. 4. THE ELECTRIC ARC MODEL The electric arc furnace is a nonlinear and unbalanced high power threephase charge. In simulations were used a model based on relations between arc radius/length, the arc voltage and the current. This model was presented in [11], [12] and used by authors in [17] and [18]. It consider that the arc voltage-current characteristic can be described by the following relationship A th ( D I ) U = U + C +. (7) In (7) U A, I A is arc voltage and current, U th is the threshold value to which voltage tends when current increases, C and D are constants whose values (C a, D a, and C b, D b ) determine the difference between the increasing and decreasing-current A

5 Modeling the electromagnetic pollution of the electric arc furnaces 169 parts of the U-I characteristic. The dynamic arc voltage-current characteristic must be an arc length function, given by relation: U A k U A0 ( I ) =. (8) In (8) U A0 represent the value of the arc voltage for a reference arc length l0 and k is the ratio between the threshold voltage value for arc length l, U th () l and the threshold voltage value for arc length l 0, U th ( l 0 ). The dynamic model for electric arc: U th = A + Bl. (9) In (9) A is a constant equal with the sum of cathode and anodic drop voltages ( A 40V ) and B represent the drop voltage on the unit length, having usual values of 10 V cm [17]. It can be obtained the dependency of k by the electric arc length: k () l ( A + B l) ( A + B ) A =. (10) l 0 5. MODELING THE ELECTRIC ARC FURNACE (EAF) INSTALLATION For model and simulate the operation of the entire installation of the threephase EAF, there are identified, the electric diagram s parameters [14, 15]; then, there are determined the parameters of the arc s model so that, further the simulation, to be obtained results very close to the results following the measurements during the electric arc s stable burning. 5.1. THE PARAMETERS OF THE ELECTRIC SCHEME COMPONENTS Parameters, resistances and inductivities, of the low voltage supply line are determined by measurements made by a specialized institute [15]. For the short network, supposed symmetric, is obtaining R p = 0.364 m Ω, L = 13.128 m Ω. (11) The total impedances values of the short network phases are given by [4]: r1 r2 r3 ( ) ( ) ( ) ( ) ( ) ( ) Z = R + 32ω M M + jω L 05. M 05. M = R + jωl, p 12 13 p 12 13 r1 r1 Z = R + 32ω M M + jω L 05. M 05. M = R + jωl, p 23 12 p 23 12 r2 r2 Z = R + 32ω M M + jω L 05. M 05. M = R + jωl. p 13 23 p 13 23 r3 r3 p (12)

170 Manuela Pănoiu, Caius Pănoiu, Ioan Şora 6 Calculation of the mutual inductivities between phases i and j can be made based on the relation [2, 4] 2 2 2 2 M = µ π + + + ij 0 2 l ln l l d ij d ij l d ij + dij, (13) where l are the phase conductors length and d ij the distance between them. The electromagnetic unbalance of the short network is due to the zone where the short network s conductors are situated in the same plan, and parallel M 12 = M 23 > M13. In case of the EAF from the analyzed installation, the length of the section where the short network conductors are situated in the same plan is l =10 m, (14) and the distance between the short netwwork s conductors d = d 1m, d = d + d 2 m. (15) 12 23 = 13 12 23 = With these values, the mutual inductivities between the phase conductors in the zone where these are in the same plan are M = M = 4.1865 H, M 13 = 2. 9853 µh. (16) 12 23 µ From the relations (11) (16) and for ω = 314 rad/s are obtaining the values of the total resistances, on each phase R = 0.6908 m Ω, R = 0.3640 m Ω, R = 0.0372 m Ω, (17) r1 r2 r3 as well as of the total inductivities L = L =.5422 µ H, L = 8.9416. (18) r 1 r3 9 r 2 µ H Because the impedances of medium voltage supply line are small compared with the ones from the low voltage line, these were included in the EAF s transformer parameters. The values of the main parameters of the EAF s transformer are 73 MVA; 30 kv/0.6 kv; /Υ/50 Hz. 5.2. CHOOSING THE VALUES OF THE MODEL S PARAMETERS The selection of the values of the model s parameters was made in such way that the wave shapes of the currents and voltages obtained following the simulation to correspond to the ones obtained following the measurements made on the real installation. In this purpose was analyzed the influence of each parameter which comes in the model and detailed present in [17], finding the following: The constants D a and D b from (7) have a small influence. The values of these constants were D a = D b = 5,000 A [11, 12, 17].

7 Modeling the electromagnetic pollution of the electric arc furnaces 171 The influence of C a and C b on the ignition voltage values is also small. Were used the values C a = 190,000 W, respectively C b = 39,000 W. In this way, for the same value of the extinction voltage the values of the ignition voltage on the two semi-alternances are different [17]. As regards the extinction voltage value used during the simulations, was found that this influences both the wave shapes and the values of currents and voltages obtained by simulation. Was admitted U th = 200 V. 6. THE SIMULATION RESULTS OF THE REAL INSTALLATION Following the measurements made on the EAF s real installation, was observed that its operation is featured by the presence of an unbalanced 3-phase regime [14]. The simulation of the EAF s operation as unbalanced 3-phase load was analyzed for two cases: unequalness of total impedances values of short network phases and unequalness of extinction voltages values on the 3 phases, due to unequalness of the electric arcs lengths on the 3 phases. In [17] was presented detailed both cases. In the first case, because there are sections of the short network where the cables of the 3 phases are disposed in the same plan, it appears an unbalance of the values of short network s total impedances, relation (12). Using these values of the short network s total impedances, as well as the parameters of the electric arc s model previously presented, was performed simulations using the diagram presented in Fig. 3. The comparisons of the simulation results with the ones of the measurements were made for the transformer operation s case, S = = 73 MVA, power which is equal with the one measured during the stable burning period after approx. 2 hours from the beginning of the heat making [14], S m = = 72.25 MVA. Are found that the value of the current s and voltage s amplitude from the low voltage line obtained by simulations corresponds to the one obtained by measurements made in the stable burning phase [14]. The spectral characteristics obtained by the measurements made on the low voltage supply line in the reduction phase correspund, from the viewpoint of the harmonics present in the specter, with the ones obtained by simulation [17]. Results that the values of the electric installation s parameters and of the chosen model s parameters allow a good reproduction of the operation of the EAF real installation. With these values of the parameters, using a Matlab program, were determined the total powers S, P, Q, D on the three phases and the power factor in deformable regime K p, as well as the values of the total harmonic distortion THDI, THDU for current and voltage on the medium voltage line, presented in Table 1.

172 Manuela Pănoiu, Caius Pănoiu, Ioan Şora 8 Fig. 3 The electric diagram for simulation using different values of the short network impedance. Table 1 Measure values Simulating values S (MVA) P (MW) Q (Mvar) D (Mvad) The power factor, K p THDI (%) THDU (%) 72.25 48.63 52.43 10.29 0.546 16.83 7.3 73.51 47.36 55.97 5.35 0.644 8.44 3.17 In case of simulation using unequal values of the electric arc s length, the influence of the unequalness of the electric arc s length on the three phases is reflecting on the different values of extinction voltage on the three phases, which determines an unbalanced regime. Following the simulations, it resulted that, using equal values of the total resistances and total inductivities of the short network s conductors, given by relation (11), choosing for the electric arc s length l = 19. 5 cm l = 16 cm l = 12. 5 cm. (19) 1 2 3 Based on relation (9), they lead to obtain the values of the electric arc s voltage given by the relation: U th1 = 235 V Uth2 = 200 V Uth3 = 165 V. (20)

9 Modeling the electromagnetic pollution of the electric arc furnaces 173 By comparing the results from Table 2, where are presented the values of the direct inverse and homopolar sequence components, of the currents and voltages resulted by simulation of the unbalance between the two situations. Following the simulations, were obtained the waveforms similar to the ones obtained in case of simulations with unequal values of the short network s impedances. Table 2 Simulation using different values of the short network phases impedance, relation (12) Simulation using different values of the arc length, relation (19) I R -458.16-j1324.82-219.22-1-j1292.5 I R 1401.81 1310.99 I S -876.95+j1248.53-875.5+j1453.2 I S 1525.73 1696.55 I T 1335.10+j76.29 1094.7-j160.7 I T 1337.28 1106.44 I + -567.17-j1300.31-575.47-j1214.3 I + 1418.62 1343.77 I - 109.48-j23.63 356.8-j77.09 I - 112.00 365.03 I 0-1e-029+j7.4e-013 4.83e-013+j4.06e-13 I 0 7.41e-013 6.3e-13 In conclusion, it is possible the same unbalance using different values of the total impedances of the short network s phases, or using different lengths of the electric arc on the three phases 7. CONCLUSIONS Following the analysis of the performed measurements results, was obtained the following important conclusions: on the low voltage supply line the voltage wave is more distorted than the current and on the medium voltage supply line the current is more distorted than the voltage; the technological phase of the heat making process has an influence on the electrice values. Following the measurements, for the medium voltage line have been obtained values for the total harmonic distortion within 8 12 % for voltage, respectively 10 40 % for current; also, have been obtained the non-symmetry in current between 10 and 60 %, respectively in voltage between 2 and 10 %. Comparing these values with the standards where the permitted total harmonic distortion s value on the medium voltage lines is of 8 %, and the average permitted value of the nonsymmetry coefficients is of 2 % [8 10], results that the installation does not matched in the international standards. It is necessary the absorbing filters for the harmonic currents and the load balancing devices of the 3-phase charge [1, 3, 18].

174 Manuela Pănoiu, Caius Pănoiu, Ioan Şora 10 Using these measurements it was performed simulations. The simulations are based on an electric arc model. The results of simulations can be used to design a complex installation for harmonics filtering, reactive power compensation and load balancing. Received on July 10, 2008 REFERENCES 1. A. Buta, A. Pană, Calitatea tensiunii Criteriu principal de analiză a interdependenţei dintre compensarea puterii reactive, echilibrarea sarcinii şi filtrarea armonicilor în reţelele de distribuţie performante, Energetica, February 1999. 2. A. Moraru, Bazele electrotehnicii. Teoria circuitelor electrice, Edit. MatrixRom, 2002. 3. A. Buta, A. Pană, Simetrizarea sarcinii reţelelor electrice de distribuţie, Edit. Orizonturi Universitare, Timişoara, 2000. 4. N. Golovanov, I. Şora, Electrotermie şi electrotehnologii, Vol. I, Edit. Tehnică, Bucureşti, 1997. 5. I. Şora, Manuela Pănoiu, C. Pănoiu, Utilizarea programului PSCAD-EMTDC la simularea funcţionării instalaţiilor cuptoarelor cu arc electric, Revista de Instrumentaţie Virtuală, Cluj, Vol. IV, n0. 1 (13), 2001. 6. T. Ionescu, O. Pop, Ingineria sistemelor de distribuţie a energiei electrice, Edit. Tehnică, Bucureşti, 1999. 7. * * *, ADA 3100/ ADA3100A, User s Manual, Real Time Devices Inc., USA, 1991. 8. * * *, Practical definitions for powers in systems with nonsinusoidal waveforms and unbalanced loads, IEEE Working Group on nonsinusoidal situations, IEEE Trans. on Power Delivery, 11, 1998, pp. 79-87. 9. * * *, Norm regulation regarding the limitation of the non-sinusoidal and non-symmetric regime in electric networks, PE 143/2001. 10. * * *, IEC 61000-4-7 Ed. 2 (2002): Electromagnetic compatibility (EMC) Part 4-7: Testing and measurement techniques General guide on harmonics and interharmonics measurements and instrumentation, for power supply systems and equipment connected thereto. 11. G. C. Montanari, M. Loggini, A. Cavallini, L. Pitti, D. Zaminelli, Arc-Furnace Model for the Study of Flicker Compensation in Electrical Networks, IEEE Transactions on Power Delivery, 9, 4, pp. 2026-2036, 1994. 12. L. Tang, S. Kolluri, F. Mark, Mc-Granaghan, Voltage Flicker Prediction for Two Simultaneously operated Arc Furnaces, IEEE Trans. on Power Delivery, 12, 2, April 1997. 13. ***, PSCAD-EMTDC, User s Manual, Manitoba HVDC Research Centre, USA, 1999. 14. Manuela Pănoiu, C. Pănoiu, I. Şora, Experimental Research Concerning the Electromagnetic Pollution Generated by the 3-Phase Electric Arc Furnaces in the Electric Power Supply Networks, Acta Electrotehnica, 47, 2, pp. 102-112, 2006. 15. G. Bălan, J. Constantinescu, B. Marcu, Ş. Cotenescu, Raţionalizarea consumului de energie electrică la cuptorul nr. 1 de 100 t pe baza bilanţului energetic, Institutul de Cercetări şi Modernizări Energetice, Bucureşti, 1983. 16. I. Chiuţă, I. Conecini, Compensarea regimului energetic deformant, Edit. Tehnică, Bucureşti, 1989. 17. Manuela Pănoiu, C. Pănoiu, I. Şora, Modeling of Three Phase Electric Arc Furnace, Acta Electrotehnica, 48, 2, pp 124-134, 2007. 18. Manuela Pănoiu, C. Pănoiu, I. Şora, Simulation results regarding harmonics filtering, reactive power compensation and load balancing on power loads, Proceedings of the 17 th IASTED International Conference on Applied Simulation and Modeling, pp. 71-76, Greece, Corfu, 2008.