- Email: [email protected]
A novel induction motor starting method using superconduction
Physica C 507 (2014) 95–102
Contents lists available at ScienceDirect
Physica C journal homepage: www.elsevier.com/locate/physc
A novel induction motor starting method using superconduction F.B.B. Silva a,c,⇑, M.T.D. Orlando b, J.F. Fardin c, D.S. Simonetti c, C.A. Baldan d a
Ifes – Federal Institute of Espírito Santo, Dept. of Industrial Automation, Serra, ES 29173087, Brazil UFES – Federal University of Espírito Santo, Dept. of Physics, Vitória, ES, Brazil c UFES – Federal University of Espírito Santo, Dept. of Electrical Engineering, Vitória, ES, Brazil d EEL/USP – Engineering School from Lorena/University of São Paulo, SP, Brazil b
a r t i c l e
i n f o
Article history: Received 4 September 2014 Accepted 23 October 2014 Available online 31 October 2014 Keywords: Three-phase induction motor Superconducting fault current limiter Superconductor tape Induction motor starting method
a b s t r a c t In this paper, an alternative method for starting up induction motors is proposed, taking into account experimental measurements. The new starting current limitation method is based on using a hightemperature superconductor. A prototype of the superconducting starting current limiter was constructed with a commercially available second-generation high-temperature superconductor YBCO tape, and this was tested with a 55-kW industrial induction motor in a 440-V/60-Hz three-phase power grid. Performance evaluations of the superconducting limiter method (applied to startup of the induction motor) were performed and were compared with a direct-on-line starter and an electronic soft starter. In addition, a computational model was developed and used for electromagnetic torque analysis of the system. As significant characteristics, our method offers the ability to limit the starting current of the induction motor with greater electromagnetic torque, reduced current waveform distortion and therefore lower harmonic pollution during startup when compared to the soft starter method. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction The use of superconducting devices such as transformers [1], motors/generators [2], transmission power cables [3] and fault current limiters [4] is promising, in light of the possibility of transporting high current values and reducing the Joule effect. In the specific case of superconducting fault current limiter (SFCL) devices, there is an instantaneous change of the device from the superconducting state to the normal state when the current drawn exceeds the critical current. This is an innovative and unique technological characteristic. Superconductors have been used successfully applied as current limiters in power systems aimed at restricting the levels of shortcircuiting of a given electrical grid [4]. The superconducting fault current limiter (SFCL) uses the critical current density value, which determines the switch over between the superconducting and resistive states, as a threshold for the high starting current. In short, above a certain high current level, usually under shortcircuit conditions, we have the resistive state, and below the normal current level, the SFCL operates in the superconducting state with zero resistance. Because the inrush current of a three-phase (3-U) induction motor (IM) is up to 10 times higher than the ⇑ Corresponding author at: Ifes – Federal Institute of Espírito Santo, Dept. of Industrial Automation, Serra, ES, Brazil. Tel.: +55 (27)3348 9234. E-mail address: fl[email protected] (F.B.B. Silva). http://dx.doi.org/10.1016/j.physc.2014.10.015 0921-4534/Ó 2014 Elsevier B.V. All rights reserved.
normal current, it is possible to use the operating principle of fault current limiters to limit the IM starting current. The conceptual analysis of a new starting current limitation for induction motors is presented in this paper, along with the first prototype of a superconductive starting current limiter (SSCL) design. Tests are carried out using a squirrel cage induction motor with 55 kW, four poles, 440 V and 60 Hz. In addition, total harmonic distortion (THD) analysis of the motor current is performed to compare the new starting method SSCL with an electronic soft starter. This is done with a focus on the power quality during the motor startup. Next, a computational simulation for the system is run to estimate the performance of the electromagnetic torque developed by the motor in both methods.
1.1. The asynchronous motor Currently, the three-phase (3-U) asynchronous or induction motor (IM) is the most commonly used electrical equipment in industry. This is because of its well-known low cost of manufacture, robust construction, reliability and low maintenance requirements. Despite its simplicity and low cost, this motor has one great drawback—a large initial current surge. Until the motor reaches its rated speed, over durations that can last for hundreds of milliseconds to a few seconds, the current remains in the range of five
96
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
to ten times the rated current. The mechanical load driven by the electric motor defines this time. In addition, because the electrical circuit is highly inductive, the peak of the current can be further increased due to the asymmetry of the current waveform [5]. In this work, we will specifically discuss the three-phase squirrel cage induction motor or simply the induction motor (IM). 1.2. Induction motor starting current The conventional way of running this type of motor is by directon-line (DOL) startup. For an induction motor, direct-on-line startup presents no problem because the IM is designed to withstand this type of startup. Although the IM is able to handle this high current during DOL startup, from the power grid point of view, the high inrush current may cause various problems that affect the electrical installation. In this context, the main drawbacks are a significant voltage drop in the supply from the power grid, electromagnetic interference in other electric equipment during operations, mechanical stress caused by the sudden increase in torque, and a need to oversize the electrical system [5]. To reduce the starting IM current, several starting current methods are available for the various types of industrial motors. Based on the literature, we can list the following [6]:
Direct-on-line starting. Wye–delta starting. Autotransformer starting. Series impedance starting. Electronic soft starter. Electronic 3-U inverter (variable-frequency drive) starting.
However, the common choices to limit the inrush current for large power asynchronous machines (typically above 37 kW) are confined to power electronic devices such as soft starters or variable-frequency drives (VFDs) [7]. Actually, the VFD is not exactly a starting current method because its main function is to control the speed of the motor. Based on common knowledge, all of the types of limitations of the inrush current during the startup of induction motors are based on reducing the applied voltage to the stator of the machine [7]. Wye–delta starting applies an initial voltage that is reduced by p a 3 factor. When the speed of the motor reaches approximately 90% of the nominal speed, the stator voltage is stabilized at the nominal voltage. The autotransformer starting method uses more voltage levels than the star-delta method, typically 50%, 65% and 80% of the rated voltage. The series impedance starting method uses an impedance to produce a drop in the applied stator voltage during the startup of the motor. Finally, the electronic soft starter uses thyristors to supply a reduced RMS voltage by switching off the voltage waveform during a variable phase angle a. As the motor speeds up, the phase angle a is decreased, resulting in a rising RMS value of the voltage. Fig. 1 shows an illustrative example of soft increase of the voltage applied to a single-phase load. 2. The superconducting starting current limiter (SSCL) 2.1. Superconducting current limiters The superconducting critical parameters are critical temperature – Tc, critical magnetic field – Hc and critical current density – Jc. For the superconductor to remain in the zero resistivity state, the temperature, magnetic field and current density values must
be below the corresponding critical values. Fig. 2 illustrates this in the form of a phase diagram. If the superconductor is inside the hatched region in Fig. 2, it will have zero electrical resistance. However, if any of the critical parameter values are exceeded, the superconductor will become resistive in behavior. Superconductors have been used successfully as current limiters in power system applications, aiming to restrict the levels of short-circuit currents in a given electrical grid [5,8]. The superconducting fault current limiting (SFCL) devices use the critical current density as a sensor to detect a high starting current and to switch between the superconducting state and the resistive state. In summary, above a high current level, usually under shortcircuit conditions, we have the resistive state, and at a normal level of current, the SCFL works in the superconducting state with zero resistance. The advances in materials engineering and the current manufacturing technologies of the second-generation high temperature superconducting (2G-HTS) tapes have enabled the use of superconductors in a wide variety of applications. Currently, 2G-HTS tapes are available with extended lengths and with a low range of variation of the critical parameters. The cost per ampere-meter ($/A m) of these tapes has decreased, enabling the construction of current limiters for many different applications [4]. 2.2. The new startup method The inrush current of a three-phase induction motor is up to 10 times higher than the nominal current. Therefore, it is possible to use the operating principle of fault current limiters to limit the starting current of an induction motor. In the superconductive starting current limiter (SSCL), if the current in the superconductor is higher than the critical current, a fast transition from the superconducting state to the normal state inserts a resistance in series with the motor impedance. This causes a voltage drop as in the series impedance method. Fig. 3 shows the proposed topology in a single-line diagram. 2.3. Design of the SSCL method The SSCL presented here was designed based on the topology shown in Fig. 4. It is made using AmSC (American Superconductor, Devens, MA) YBCO Coated Conductor (CC) tape with the following specifications: width 4.4 mm, thickness 0.29 mm, measured critical current 77 A, and linear resistance per unit length (@ 300 K) 0.32 X/m. The starting time of the motor can last a few seconds. However, YBCO CC tape cannot withstand high currents over any significant length of time. Therefore, it is necessary to disconnect the superconductor tape and drive the current to a shunt reactor. We used a microcontroller timer to control the opening of a 3-U solid-state relay (an electronic switch) that turns off the current to the tape. The superconductor tape needs to turn off before its temperature reaches 300 K. To accomplish this, the shunt reactor will take on the starting current after 80 ms. This procedure avoids damage to the superconductor [9–11]. The shunt reactor consists of a 3-U air core inductor with 1 mH and 0.05 X per phase. The reactor is able to withstand the starting current during the time necessary for the motor to reach its nominal speed. The aim of the new method is to limit the maximum value of the inrush current to half of the DOL current. Therefore, the magnitude of the impedance of the SSCL (|ZSSCL|) must be greater than or equal to the magnitude of the impedance of the motor (|Zsmot|) during startup. Based on the motor data, for a 55-kW, 440-V, 60Hz, and 3-U IM, the rms value of the DOL current is Ismot = 650 A; then, |Zsmot| = 0.39 X and |ZSSCL| P 0.39 X.
97
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
Fig. 1. Voltage waveform from soft starter.
J NORMAL STATE
JC T HC
H
TC SUPERCONDUCTING STATE
Fig. 2. Phase diagram and critical parameters.
Starting Current Limiter Shunt Reactor
stabilizer to ensure homogeneous transition along the entire length of the tape. In accordance with project specifications, the phase voltage is split in half between |Zsmot| and |ZSSCL|. The electric field developed within the tape must be lower than 0.5 V/cm, and the peak value of the voltage applied to the SSCL must be 180 V. Hence, the minimum required length of the tape is 3.6 m. The support mounting of the SSCL was made with a G10 epoxy fiberglass cylindrical tube 19 cm in diameter and 13 cm in height. The tape was bifilarly wound to cancel the inductive effect of the SSCL. For tape protection, we used a shunt resistor wound on the inner surface of the G10 cylindrical support. The shunt resistivity was set to 0.175 X/m at 300 K, so that the total resistance at 300 K of the arrangement (Fig. 5) was 0.5 X using 5.35 m of tape and 4 m of shunt. These values were chosen so that when a complete transition to the resistive state occurs in the SSCL, the total inrush starting current will be lowered to half of the DOL starting motor current. This prototype is designed to operate immersed in a liquid nitrogen bath. Fig. 4 shows a single phase of the superconductor device that was used to build the SSCL experimental device. 3. Experimental results of the new method
Superconductor
Fig. 3. Simplified single-line diagram of the SSCL method.
To validate the expected capability of the SSCL to withstand the starting current of a real industrial motor, a test was performed with the 55-kW, 440-V, 3-U squirrel cage motor. First, the limiting effects of the SSCL were confirmed by comparison with the DOL starting current waveforms for the A phase. Then, harmonic analyses were carried out to verify and compare the THD of the soft starter method and the SSCL method.
Fig. 4. One phase of the SSCL device.
To ensure safe operation of the superconductor tape, the device has a resistive shunt protector connected every 20 cm along the length of the tape. Furthermore, at each shunt connection point on the tape, a copper bar was added. The bar functions as a thermal
Direct-on-line starting Current (A)
1500
1240 A
1000
500
0
-500
-1000 0,0
0,2
0,4
0,6
Time (s) Fig. 5. Measured DOL starting current.
0,8
1,0
98
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
3.1. Material and methods
1500
3.2. Analysis of the current limiting A comparison between the DOL and SSCL starting current waveforms for the 55-kW motor is shown in Figs. 5 and 6. The DOL starting current has an initial peak of 1240 A and a peak of 750 A after five cycles. For the SSCL method, the starting motor current was limited to 700 A for the first peak and 412 A after 5 cycles. The reduction of the starting current was approximately 45%, slightly below the design specification of 50%. This suggests that a complete transition of the tape did not occur, and the final temperature of 300 K was not reached. The experimental results for the new method showed a limitation of the starting current that was within the expected range and in accordance with the design specifications.
700 A
SSCL Current (A)
1000 500 0 -500 -1000 -1500 0,0
0,2
0,4
0,6
0,8
Fig. 6. Measured SSCL starting current.
Fig. 7 shows the envelope obtained from an overview of the current waveform of the soft starter, and Fig. 8 shows a zoom-in view of the first two cycles of the (a) SSCL current waveform and (b) soft starter current waveform. The equivalent impedance of the SSCL results in a more linear electric circuit than the electronic circuit imposed by the soft starter. The distortion of the waveform will result in a high percentage
400
200
0
-200
-400 0
1
2
Time (s) Fig. 7. Soft starter current envelope.
3.3. Analysis of the THD A typical criterion for evaluating the power quality of electric systems is total harmonic distortion (THD). The THD of a waveform is defined as the ratio of the sum of the powers of all harmonic components to the power of the fundamental frequency. Harmonic analysis was performed on the soft starter current and the SSCL current. Data acquisition with the power analyzer generated the results that are plotted in this subsection.
Table 1 Nameplate data of the IM tested. Rated horsepower Rated voltage Rated full-load amperage Power factor Full-load efficiency Rated full-load speed Ismot/Irated
1,0
Time (s)
Soft-Starter Current (A)
For the experiments reported here, we used a differential 3-U power analyzer to measure the power and the THD during the motor startup in each case. All of the experimental THD analyses were carried out with a sample time of 100 ms. The data acquisition for the voltage and current waveforms was performed using a 100 MHz – 1 GS/s four-isolated channel digital storage oscilloscope equipped with high voltage differential probes and AC current probes. The motor nameplate data used in the experiments are listed in Table 1. We used an electronic soft starter device with six Siliconcontrolled Rectifiers (SCRs). The phase angle power controller had the following specifications: voltage line input = 220 up to 575 V; voltage output = 220 up to 575 V; rated output current = 130 A; maximum output current = 650 A. This device was configured for testing, and the main settings used were a control for a rising voltage ramp, an initial phase-to-phase voltage of 110 V and an acceleration time of 2 s. In accordance with the requirements of our project, the prototype that we built is able to withstand the DOL starting current of the motor. All of the tests were performed with no mechanical load during the operation of the motor. Initially, the SSCL method was compared with the DOL method to verify the ability to limit the current. Later, a comparison was performed between the SSCL method and the soft starter method to validate the computational results and to measure the THD. All of the tests were performed in the laboratory of industrial electric motors of the mining company Vale SA.
75 HP (55 KW) 440 V 87.5 A 0.88 Lagging 91.9 1770 RPM 7.4 Fig. 8. Current waveform (a) SSCL and (b) soft starter.
99
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
40 3rd Order 5th Order 7th Order 9th Order 11th Order
35 30
Harmonic (%)
of harmonics and a higher value of the THD in the soft starter method. Fig. 9 shows the results of the harmonic analysis for the soft starter current for the A phase of the IM, and Fig. 10 shows the resulting THD. It can be observed from Fig. 9 that the 5th harmonic is predominant. Fig. 11 shows the results of the harmonic analysis of the current for the A phase of the IM with the SSCL limiting the starting current. The THD for this method was calculated and is shown in Fig. 12. Again, the 5th harmonic is predominant but at a lower value than when using the soft starter. To study the torque developed by the new method, a torque transducer that can operate at the nominal rated IM power during the startup transient is required. Unfortunately, none of the laboratories that we consulted had the necessary equipment. Thus, we decided to obtain an estimate of these results by computer simulation, as described below.
25 20 15 10 5 0 0
1
2
3
Time (s) Fig. 11. Harmonic components of the SSCL current.
4. Computational simulation and comparison between the SSCL and the electronic soft starter
40 35
In this section, we used a computational simulation to evaluate the electromagnetic torque provided by the SSCL and the soft starter and to compare the performance of the two starting current
THD (%)
40 3rd Order 5th Order 7th Order 9th Order 11th Order
35 30
Harmonic (%)
30 25 20 15 4.25 %
10 5
25
0 0
20
1
2
3
Time (s) 15 Fig. 12. THD of the soft starter.
10 5 0 0
1
2
3
Time (s) Fig. 9. Harmonic components of the soft starter.
40
36 %
methods. All of the simulation results were obtained using Matlab/Simulink. The model of the IM was set with the values of the parameters obtained from the blocked rotor test and from the no-load test. The model of the 3-U soft starter was built using six thyristors fired by a pulse generator module. The set point to the pulse generator is a ramp with a rise time of approximately 2 s. The model of the system SSCL was built using a variable resistor in parallel with a shunt reactor, and their values were set with approximately the same values as those of the real elements.
35
4.1. Setting the model
THD (%)
30 25 20 15 10 5 0 0
1
Time (s) Fig. 10. THD of the soft starter.
2
3
The electronic soft starter is a non-linear variable voltage source that supplies limited but distorted current waveforms to the motor during startup, as observed in Fig. 1. Consequently, the main purpose of the current limiting is achieved, but harmonic distortion of current waveform is accentuated. For example, the real device used in the experimental tests has an increasing current ramp type, and its current waveform results in greater distortion, as can be observed in Fig. 13. Lower current amplitudes cause further distortion in the waveform [12]. More details can be seen in Fig. 14. The soft starter model is representative of the system. However, there are some differences between the model and the real device, for example, with the current control circuit and the harmonic
100
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
90
400
Soft Starter Current (A)
Soft Starter Current (A)
600
200 0 -200 -400
0
-90 -600 0,0
0,5
1,0
1,5
2,0
2,5 0,10
Time (s) Fig. 13. Simulated soft starter ramp current control.
0,14
Fig. 14. Zoom at the soft starter current.
1000
1500 1000
Soft Starter Current (A)
Soft Starter Current (A)
0,12
Time (s)
500 0 -500 -1000
500
0
-500
-1500 0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
-1000
Time (s)
0,04
0,06
0,08
Time (s)
Fig. 15. Simulated soft starter constant current control.
Fig. 16. Zoom at the soft starter current.
1500
1000
500
500
SSCL Current (A)
SSCL Current (A)
1000
0 -500 -1000 -1500 0,0
0
-500
0,2
0,4
0,6
0,8
1,0
1,2
1,4
Time (s) Fig. 17. Simulated SSCL starting current.
-1000 0,00
0,02
0,04
0,06
Time (s) Fig. 18. Zoom at the SSCL starting current.
filter circuit. These circuits were not included in the simulation, resulting in a small difference in outcomes. For a fair comparison of the torque, the soft start model was adjusted to provide a supply of current to the IM that was similar to that of the SSCL method. In Fig. 15, the envelope of the waveform of the line current to the IM indicates a soft starter condition that is set to be similar to the SSCL method. Next, in Fig. 16, we can see a zoomed-in view of the first cycles of the current. This view
shows a clear presence of distortion. In fact, this current waveform indicates the presence of high THD, in accordance with the experimental results. The SSCL method produces an approximately linear voltage drop due to resistive behavior of the superconductor tape. Therefore, the harmonic distortion of the current will be lower in this case than in the soft starter method.
101
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
We can view both the envelope and waveform of the SSCL current obtained by simulation in Figs. 17 and 18.
400
4.2. Torque analysis
300
Direct-on-line starting current (A)
1500
1232 A
1000
SSCL Torque (N.m)
200
100
0 1,0
0,8
0,6
0,4
0,2
0,0
Slip (pu) Fig. 21. SSCL electromagnetic torque.
400
Soft Starter Torque (N.m)
We can use the computational model to obtain the electromagnetic (EM) torque behavior during the starting of the IM using the soft starter and the SSCL methods. Electromagnetic torque is the torque that corresponds to the power transferred across the air gap. This is not the same as the output torque, or the load torque, that is available at the shaft because it does not account for friction, windage and stray load losses. It is well known that the EM torque (for any type of method) is quite oscillatory because the electromagnetic dynamics are faster than the mechanical dynamics. However, for the two starting methods presented in this paper, it is possible to compare one to the other. The current control mode of the soft starter has been set as a constant current mode during the startup of the IM. This was done to enable comparison of the two methods under similar conditions. In both methods, the current was limited to about half of the simulated DOL starting current (peak value). This can be observed in Fig. 19. Figs. 20–22 present the graphs of the DOL, SSCL and soft starter EM torque, respectively.
341 N.m
300 200 N.m
200
100
500 0
0
1,0
0,8
0,6
0,4
0,2
0,0
Slip (pu)
-500
Fig. 22. Soft starter electromagnetic torque.
-1000 -1500 -0,1 0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
Time (s)
The DOL maximum EM torque corresponds to approximately two times the SSCL maximum EM torque. The soft starter EM torque has plenty of noise from high frequency harmonic components of its current. For the soft starter, the high distortion of the current waveform results in an EM torque that is approximately 40% lower than the SSCL torque, for the same limited peak current. The models used in the simulations showed results that are consistent with our expectations. This is evidenced in the experimental tests and results described above.
Fig. 19. Simulated DOL starting current.
691 N.m
800
DOL Torque (N.m)
Some aspects should be highlighted for a proper comparison of the EM torque produced in each case.
600
5. Conclusions 400
200
0 1,0
0,8
0,6
0,4
Slip (pu) Fig. 20. DOL electromagnetic torque.
0,2
0,0
The performance of the new SSCL method confirmed the main goal of limiting the starting current of the IM. In experimental tests, a comparison with the DOL starting method showed a reduction of approximately 45%, which is close to the 50% value of the design specification. This slightly lower reduction indicates that the tape temperature was below 300 K, indicating safe operation of the superconductor module. THD measurement during the startup showed a maximum value of 36% for the soft starter, and a maximum value of 4.25% for the SSCL method. The harmonic pollution produced by the soft
102
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
starter was neglected due to the short starting time. Altogether, in specific cases of industries where multiple IMs of large capacity are frequently started, and where the power systems have strict requirements for power quality indices, this high THD value of soft starters is a problem that must be addressed [13]. The EM torque produced in each method was investigated by simulation. The DOL maximum EM torque corresponds to about two times the SSCL maximum EM torque, and the soft starter method results in an EM torque that is approximately 40% lower than the SSCL torque. The models used in the simulations showed results that are consistent with our expectations. Based on simulation and experiments, we noted that the SSCL current is clearly less distorted than the soft starter current. In comparison, the superconductor starting current method offers a lower distortion of the current waveform, resulting in lower harmonic pollution during startup. In addition, we highlight other advantages of the SSCL such as greater robustness and a smaller number of electronic components. However, the SSCL method (using a YBCO 2G CC tape) does have some disadvantages. These include inhomogeneity of the superconducting tape, AC losses, losses by thermal conduction along the current leads, and high cost of operation with a cryogenic system required to maintain the superconductor at temperatures below its critical temperature [14,15]. The next step of this work is to test additional starting methods under different load conditions and under steady-state operation. Acknowledgements This work was supported by FAPES-PRONEX 48508675/2009 and CNPq-Universal 54694825/2011. The experimental results were obtained in the laboratory of electric motors in the Tubarão Complex/VALE SA with the help of the electrical engineer Flavio Helmer Ferro.
References [1] T. Hoshino, I. Muta, Load test on superconducting transformer and fault current limiting devices for electric power system, IEEE Trans. Magnetics 30 (4) (1994) 2018–2021. [2] D.U. Gubser, Superconducting motors and generators for naval applications, Physica C 392–396 (2) (2003) 1192–1195. [3] N. Amemiya, Q. Li, K. Takeuchi, T. Nakamura, M. Yagi, S. Mukoyama, Y. Aoki, N. Fujiwara, Effects of lateral-tailoring of coated conductor for AC loss reduction of superconducting power transmission cables, IEEE Trans. Appl. Supercond. 21 (3) (2011) 943–946. [4] M. Noe, M. e Steurer, High-temperature superconductor fault current limiters: concepts, applications and development status, Supercond. Sci. Technol. 20 (3) (2007) R15–R29. [5] Y. Hu, Z. Chen, Z. Chen, Y. Yuan, A novel method for the analysis of the impact of the high-power motor starting on grid, in: 4th International Conference on Electric Utility Deregulation and Restructuring and Power Technologies. IEEE, 2011, pp. 394–398. [6] C. Sen, Principles of Electric Machines and Power Electronics, third ed., John Wiley & Sons, New York, 2013. [7] L. Ghodhbani, B. Rebhi, F. Kourda, M. Elleuch, M. Abidi, Comparison among electronic start up methods for induction motors, in: IEEE 7th Int. Multi-Conf. on Syst. Signals and Devices, 2010, pp. 1–7. [8] A. Morandi, Fault current limiter: an enabler for increasing safety and power quality of distribution networks, IEEE Trans. Appl. Supercond. 23 (2013) 5604608. [9] C. Gandioli, P. Tixador, G.B. Mariani, Tests and simulations of different YBCO tape for FCL, IEEE Trans. Appl. Supercond. 22 (2012) 56003104. [10] W. Schmidt, H.-P. Kraemer, H.-W. Neumueller, U. Schoop, D. Verebelyi, A.P. Malozemoff, Investigation of YBCO coated conductors for fault current limiter applications, IEEE Trans. App. Supercond. 17 (2) (2007) 3471–3474. [11] C. Schacherer et al., Dissipated energy as a design parameter of coated conductor for their use in resistive fault current limiters, J. Phys: Conf. Ser. 97 (2008) 012193. [12] A. Riyaz, A. Iqbal, S. Moinoddin, S.M. Ahmed, H. Abu-Hub, Comparative performance analysis of thyristor and IGBT based induction motor soft starters, Int. J. Eng., Sci. Technol. 1 (1) (2009) 90–105. [13] G. Bhuvaneswari, S. Charles, M.G. Nair, Power quality studies on a soft-start for an induction motor, Transmission and Distribution Conference and Exposition, T&D. IEEE/PES, 2008, pp. 1–6. [14] S. Imparato et al., Experimental evaluation of AC losses of a DC restive SFCL prototype, IEEE Trans. Appl. Supercond. 20 (3) (2010) 1199–1202. [15] L. Ying, J. Sheng, B. Lin, L. Yao, J. Zhang, Z. Jin, Y. Li, Z. Hong, AC loss and contact resistance of resistive type fault current limiter using YBCO coated conductors, IEEE Trans. Appl. Supercond. 22 (2012) 5602404.
Contents lists available at ScienceDirect
Physica C journal homepage: www.elsevier.com/locate/physc
A novel induction motor starting method using superconduction F.B.B. Silva a,c,⇑, M.T.D. Orlando b, J.F. Fardin c, D.S. Simonetti c, C.A. Baldan d a
Ifes – Federal Institute of Espírito Santo, Dept. of Industrial Automation, Serra, ES 29173087, Brazil UFES – Federal University of Espírito Santo, Dept. of Physics, Vitória, ES, Brazil c UFES – Federal University of Espírito Santo, Dept. of Electrical Engineering, Vitória, ES, Brazil d EEL/USP – Engineering School from Lorena/University of São Paulo, SP, Brazil b
a r t i c l e
i n f o
Article history: Received 4 September 2014 Accepted 23 October 2014 Available online 31 October 2014 Keywords: Three-phase induction motor Superconducting fault current limiter Superconductor tape Induction motor starting method
a b s t r a c t In this paper, an alternative method for starting up induction motors is proposed, taking into account experimental measurements. The new starting current limitation method is based on using a hightemperature superconductor. A prototype of the superconducting starting current limiter was constructed with a commercially available second-generation high-temperature superconductor YBCO tape, and this was tested with a 55-kW industrial induction motor in a 440-V/60-Hz three-phase power grid. Performance evaluations of the superconducting limiter method (applied to startup of the induction motor) were performed and were compared with a direct-on-line starter and an electronic soft starter. In addition, a computational model was developed and used for electromagnetic torque analysis of the system. As significant characteristics, our method offers the ability to limit the starting current of the induction motor with greater electromagnetic torque, reduced current waveform distortion and therefore lower harmonic pollution during startup when compared to the soft starter method. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction The use of superconducting devices such as transformers [1], motors/generators [2], transmission power cables [3] and fault current limiters [4] is promising, in light of the possibility of transporting high current values and reducing the Joule effect. In the specific case of superconducting fault current limiter (SFCL) devices, there is an instantaneous change of the device from the superconducting state to the normal state when the current drawn exceeds the critical current. This is an innovative and unique technological characteristic. Superconductors have been used successfully applied as current limiters in power systems aimed at restricting the levels of shortcircuiting of a given electrical grid [4]. The superconducting fault current limiter (SFCL) uses the critical current density value, which determines the switch over between the superconducting and resistive states, as a threshold for the high starting current. In short, above a certain high current level, usually under shortcircuit conditions, we have the resistive state, and below the normal current level, the SFCL operates in the superconducting state with zero resistance. Because the inrush current of a three-phase (3-U) induction motor (IM) is up to 10 times higher than the ⇑ Corresponding author at: Ifes – Federal Institute of Espírito Santo, Dept. of Industrial Automation, Serra, ES, Brazil. Tel.: +55 (27)3348 9234. E-mail address: fl[email protected] (F.B.B. Silva). http://dx.doi.org/10.1016/j.physc.2014.10.015 0921-4534/Ó 2014 Elsevier B.V. All rights reserved.
normal current, it is possible to use the operating principle of fault current limiters to limit the IM starting current. The conceptual analysis of a new starting current limitation for induction motors is presented in this paper, along with the first prototype of a superconductive starting current limiter (SSCL) design. Tests are carried out using a squirrel cage induction motor with 55 kW, four poles, 440 V and 60 Hz. In addition, total harmonic distortion (THD) analysis of the motor current is performed to compare the new starting method SSCL with an electronic soft starter. This is done with a focus on the power quality during the motor startup. Next, a computational simulation for the system is run to estimate the performance of the electromagnetic torque developed by the motor in both methods.
1.1. The asynchronous motor Currently, the three-phase (3-U) asynchronous or induction motor (IM) is the most commonly used electrical equipment in industry. This is because of its well-known low cost of manufacture, robust construction, reliability and low maintenance requirements. Despite its simplicity and low cost, this motor has one great drawback—a large initial current surge. Until the motor reaches its rated speed, over durations that can last for hundreds of milliseconds to a few seconds, the current remains in the range of five
96
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
to ten times the rated current. The mechanical load driven by the electric motor defines this time. In addition, because the electrical circuit is highly inductive, the peak of the current can be further increased due to the asymmetry of the current waveform [5]. In this work, we will specifically discuss the three-phase squirrel cage induction motor or simply the induction motor (IM). 1.2. Induction motor starting current The conventional way of running this type of motor is by directon-line (DOL) startup. For an induction motor, direct-on-line startup presents no problem because the IM is designed to withstand this type of startup. Although the IM is able to handle this high current during DOL startup, from the power grid point of view, the high inrush current may cause various problems that affect the electrical installation. In this context, the main drawbacks are a significant voltage drop in the supply from the power grid, electromagnetic interference in other electric equipment during operations, mechanical stress caused by the sudden increase in torque, and a need to oversize the electrical system [5]. To reduce the starting IM current, several starting current methods are available for the various types of industrial motors. Based on the literature, we can list the following [6]:
Direct-on-line starting. Wye–delta starting. Autotransformer starting. Series impedance starting. Electronic soft starter. Electronic 3-U inverter (variable-frequency drive) starting.
However, the common choices to limit the inrush current for large power asynchronous machines (typically above 37 kW) are confined to power electronic devices such as soft starters or variable-frequency drives (VFDs) [7]. Actually, the VFD is not exactly a starting current method because its main function is to control the speed of the motor. Based on common knowledge, all of the types of limitations of the inrush current during the startup of induction motors are based on reducing the applied voltage to the stator of the machine [7]. Wye–delta starting applies an initial voltage that is reduced by p a 3 factor. When the speed of the motor reaches approximately 90% of the nominal speed, the stator voltage is stabilized at the nominal voltage. The autotransformer starting method uses more voltage levels than the star-delta method, typically 50%, 65% and 80% of the rated voltage. The series impedance starting method uses an impedance to produce a drop in the applied stator voltage during the startup of the motor. Finally, the electronic soft starter uses thyristors to supply a reduced RMS voltage by switching off the voltage waveform during a variable phase angle a. As the motor speeds up, the phase angle a is decreased, resulting in a rising RMS value of the voltage. Fig. 1 shows an illustrative example of soft increase of the voltage applied to a single-phase load. 2. The superconducting starting current limiter (SSCL) 2.1. Superconducting current limiters The superconducting critical parameters are critical temperature – Tc, critical magnetic field – Hc and critical current density – Jc. For the superconductor to remain in the zero resistivity state, the temperature, magnetic field and current density values must
be below the corresponding critical values. Fig. 2 illustrates this in the form of a phase diagram. If the superconductor is inside the hatched region in Fig. 2, it will have zero electrical resistance. However, if any of the critical parameter values are exceeded, the superconductor will become resistive in behavior. Superconductors have been used successfully as current limiters in power system applications, aiming to restrict the levels of short-circuit currents in a given electrical grid [5,8]. The superconducting fault current limiting (SFCL) devices use the critical current density as a sensor to detect a high starting current and to switch between the superconducting state and the resistive state. In summary, above a high current level, usually under shortcircuit conditions, we have the resistive state, and at a normal level of current, the SCFL works in the superconducting state with zero resistance. The advances in materials engineering and the current manufacturing technologies of the second-generation high temperature superconducting (2G-HTS) tapes have enabled the use of superconductors in a wide variety of applications. Currently, 2G-HTS tapes are available with extended lengths and with a low range of variation of the critical parameters. The cost per ampere-meter ($/A m) of these tapes has decreased, enabling the construction of current limiters for many different applications [4]. 2.2. The new startup method The inrush current of a three-phase induction motor is up to 10 times higher than the nominal current. Therefore, it is possible to use the operating principle of fault current limiters to limit the starting current of an induction motor. In the superconductive starting current limiter (SSCL), if the current in the superconductor is higher than the critical current, a fast transition from the superconducting state to the normal state inserts a resistance in series with the motor impedance. This causes a voltage drop as in the series impedance method. Fig. 3 shows the proposed topology in a single-line diagram. 2.3. Design of the SSCL method The SSCL presented here was designed based on the topology shown in Fig. 4. It is made using AmSC (American Superconductor, Devens, MA) YBCO Coated Conductor (CC) tape with the following specifications: width 4.4 mm, thickness 0.29 mm, measured critical current 77 A, and linear resistance per unit length (@ 300 K) 0.32 X/m. The starting time of the motor can last a few seconds. However, YBCO CC tape cannot withstand high currents over any significant length of time. Therefore, it is necessary to disconnect the superconductor tape and drive the current to a shunt reactor. We used a microcontroller timer to control the opening of a 3-U solid-state relay (an electronic switch) that turns off the current to the tape. The superconductor tape needs to turn off before its temperature reaches 300 K. To accomplish this, the shunt reactor will take on the starting current after 80 ms. This procedure avoids damage to the superconductor [9–11]. The shunt reactor consists of a 3-U air core inductor with 1 mH and 0.05 X per phase. The reactor is able to withstand the starting current during the time necessary for the motor to reach its nominal speed. The aim of the new method is to limit the maximum value of the inrush current to half of the DOL current. Therefore, the magnitude of the impedance of the SSCL (|ZSSCL|) must be greater than or equal to the magnitude of the impedance of the motor (|Zsmot|) during startup. Based on the motor data, for a 55-kW, 440-V, 60Hz, and 3-U IM, the rms value of the DOL current is Ismot = 650 A; then, |Zsmot| = 0.39 X and |ZSSCL| P 0.39 X.
97
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
Fig. 1. Voltage waveform from soft starter.
J NORMAL STATE
JC T HC
H
TC SUPERCONDUCTING STATE
Fig. 2. Phase diagram and critical parameters.
Starting Current Limiter Shunt Reactor
stabilizer to ensure homogeneous transition along the entire length of the tape. In accordance with project specifications, the phase voltage is split in half between |Zsmot| and |ZSSCL|. The electric field developed within the tape must be lower than 0.5 V/cm, and the peak value of the voltage applied to the SSCL must be 180 V. Hence, the minimum required length of the tape is 3.6 m. The support mounting of the SSCL was made with a G10 epoxy fiberglass cylindrical tube 19 cm in diameter and 13 cm in height. The tape was bifilarly wound to cancel the inductive effect of the SSCL. For tape protection, we used a shunt resistor wound on the inner surface of the G10 cylindrical support. The shunt resistivity was set to 0.175 X/m at 300 K, so that the total resistance at 300 K of the arrangement (Fig. 5) was 0.5 X using 5.35 m of tape and 4 m of shunt. These values were chosen so that when a complete transition to the resistive state occurs in the SSCL, the total inrush starting current will be lowered to half of the DOL starting motor current. This prototype is designed to operate immersed in a liquid nitrogen bath. Fig. 4 shows a single phase of the superconductor device that was used to build the SSCL experimental device. 3. Experimental results of the new method
Superconductor
Fig. 3. Simplified single-line diagram of the SSCL method.
To validate the expected capability of the SSCL to withstand the starting current of a real industrial motor, a test was performed with the 55-kW, 440-V, 3-U squirrel cage motor. First, the limiting effects of the SSCL were confirmed by comparison with the DOL starting current waveforms for the A phase. Then, harmonic analyses were carried out to verify and compare the THD of the soft starter method and the SSCL method.
Fig. 4. One phase of the SSCL device.
To ensure safe operation of the superconductor tape, the device has a resistive shunt protector connected every 20 cm along the length of the tape. Furthermore, at each shunt connection point on the tape, a copper bar was added. The bar functions as a thermal
Direct-on-line starting Current (A)
1500
1240 A
1000
500
0
-500
-1000 0,0
0,2
0,4
0,6
Time (s) Fig. 5. Measured DOL starting current.
0,8
1,0
98
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
3.1. Material and methods
1500
3.2. Analysis of the current limiting A comparison between the DOL and SSCL starting current waveforms for the 55-kW motor is shown in Figs. 5 and 6. The DOL starting current has an initial peak of 1240 A and a peak of 750 A after five cycles. For the SSCL method, the starting motor current was limited to 700 A for the first peak and 412 A after 5 cycles. The reduction of the starting current was approximately 45%, slightly below the design specification of 50%. This suggests that a complete transition of the tape did not occur, and the final temperature of 300 K was not reached. The experimental results for the new method showed a limitation of the starting current that was within the expected range and in accordance with the design specifications.
700 A
SSCL Current (A)
1000 500 0 -500 -1000 -1500 0,0
0,2
0,4
0,6
0,8
Fig. 6. Measured SSCL starting current.
Fig. 7 shows the envelope obtained from an overview of the current waveform of the soft starter, and Fig. 8 shows a zoom-in view of the first two cycles of the (a) SSCL current waveform and (b) soft starter current waveform. The equivalent impedance of the SSCL results in a more linear electric circuit than the electronic circuit imposed by the soft starter. The distortion of the waveform will result in a high percentage
400
200
0
-200
-400 0
1
2
Time (s) Fig. 7. Soft starter current envelope.
3.3. Analysis of the THD A typical criterion for evaluating the power quality of electric systems is total harmonic distortion (THD). The THD of a waveform is defined as the ratio of the sum of the powers of all harmonic components to the power of the fundamental frequency. Harmonic analysis was performed on the soft starter current and the SSCL current. Data acquisition with the power analyzer generated the results that are plotted in this subsection.
Table 1 Nameplate data of the IM tested. Rated horsepower Rated voltage Rated full-load amperage Power factor Full-load efficiency Rated full-load speed Ismot/Irated
1,0
Time (s)
Soft-Starter Current (A)
For the experiments reported here, we used a differential 3-U power analyzer to measure the power and the THD during the motor startup in each case. All of the experimental THD analyses were carried out with a sample time of 100 ms. The data acquisition for the voltage and current waveforms was performed using a 100 MHz – 1 GS/s four-isolated channel digital storage oscilloscope equipped with high voltage differential probes and AC current probes. The motor nameplate data used in the experiments are listed in Table 1. We used an electronic soft starter device with six Siliconcontrolled Rectifiers (SCRs). The phase angle power controller had the following specifications: voltage line input = 220 up to 575 V; voltage output = 220 up to 575 V; rated output current = 130 A; maximum output current = 650 A. This device was configured for testing, and the main settings used were a control for a rising voltage ramp, an initial phase-to-phase voltage of 110 V and an acceleration time of 2 s. In accordance with the requirements of our project, the prototype that we built is able to withstand the DOL starting current of the motor. All of the tests were performed with no mechanical load during the operation of the motor. Initially, the SSCL method was compared with the DOL method to verify the ability to limit the current. Later, a comparison was performed between the SSCL method and the soft starter method to validate the computational results and to measure the THD. All of the tests were performed in the laboratory of industrial electric motors of the mining company Vale SA.
75 HP (55 KW) 440 V 87.5 A 0.88 Lagging 91.9 1770 RPM 7.4 Fig. 8. Current waveform (a) SSCL and (b) soft starter.
99
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
40 3rd Order 5th Order 7th Order 9th Order 11th Order
35 30
Harmonic (%)
of harmonics and a higher value of the THD in the soft starter method. Fig. 9 shows the results of the harmonic analysis for the soft starter current for the A phase of the IM, and Fig. 10 shows the resulting THD. It can be observed from Fig. 9 that the 5th harmonic is predominant. Fig. 11 shows the results of the harmonic analysis of the current for the A phase of the IM with the SSCL limiting the starting current. The THD for this method was calculated and is shown in Fig. 12. Again, the 5th harmonic is predominant but at a lower value than when using the soft starter. To study the torque developed by the new method, a torque transducer that can operate at the nominal rated IM power during the startup transient is required. Unfortunately, none of the laboratories that we consulted had the necessary equipment. Thus, we decided to obtain an estimate of these results by computer simulation, as described below.
25 20 15 10 5 0 0
1
2
3
Time (s) Fig. 11. Harmonic components of the SSCL current.
4. Computational simulation and comparison between the SSCL and the electronic soft starter
40 35
In this section, we used a computational simulation to evaluate the electromagnetic torque provided by the SSCL and the soft starter and to compare the performance of the two starting current
THD (%)
40 3rd Order 5th Order 7th Order 9th Order 11th Order
35 30
Harmonic (%)
30 25 20 15 4.25 %
10 5
25
0 0
20
1
2
3
Time (s) 15 Fig. 12. THD of the soft starter.
10 5 0 0
1
2
3
Time (s) Fig. 9. Harmonic components of the soft starter.
40
36 %
methods. All of the simulation results were obtained using Matlab/Simulink. The model of the IM was set with the values of the parameters obtained from the blocked rotor test and from the no-load test. The model of the 3-U soft starter was built using six thyristors fired by a pulse generator module. The set point to the pulse generator is a ramp with a rise time of approximately 2 s. The model of the system SSCL was built using a variable resistor in parallel with a shunt reactor, and their values were set with approximately the same values as those of the real elements.
35
4.1. Setting the model
THD (%)
30 25 20 15 10 5 0 0
1
Time (s) Fig. 10. THD of the soft starter.
2
3
The electronic soft starter is a non-linear variable voltage source that supplies limited but distorted current waveforms to the motor during startup, as observed in Fig. 1. Consequently, the main purpose of the current limiting is achieved, but harmonic distortion of current waveform is accentuated. For example, the real device used in the experimental tests has an increasing current ramp type, and its current waveform results in greater distortion, as can be observed in Fig. 13. Lower current amplitudes cause further distortion in the waveform [12]. More details can be seen in Fig. 14. The soft starter model is representative of the system. However, there are some differences between the model and the real device, for example, with the current control circuit and the harmonic
100
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
90
400
Soft Starter Current (A)
Soft Starter Current (A)
600
200 0 -200 -400
0
-90 -600 0,0
0,5
1,0
1,5
2,0
2,5 0,10
Time (s) Fig. 13. Simulated soft starter ramp current control.
0,14
Fig. 14. Zoom at the soft starter current.
1000
1500 1000
Soft Starter Current (A)
Soft Starter Current (A)
0,12
Time (s)
500 0 -500 -1000
500
0
-500
-1500 0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
-1000
Time (s)
0,04
0,06
0,08
Time (s)
Fig. 15. Simulated soft starter constant current control.
Fig. 16. Zoom at the soft starter current.
1500
1000
500
500
SSCL Current (A)
SSCL Current (A)
1000
0 -500 -1000 -1500 0,0
0
-500
0,2
0,4
0,6
0,8
1,0
1,2
1,4
Time (s) Fig. 17. Simulated SSCL starting current.
-1000 0,00
0,02
0,04
0,06
Time (s) Fig. 18. Zoom at the SSCL starting current.
filter circuit. These circuits were not included in the simulation, resulting in a small difference in outcomes. For a fair comparison of the torque, the soft start model was adjusted to provide a supply of current to the IM that was similar to that of the SSCL method. In Fig. 15, the envelope of the waveform of the line current to the IM indicates a soft starter condition that is set to be similar to the SSCL method. Next, in Fig. 16, we can see a zoomed-in view of the first cycles of the current. This view
shows a clear presence of distortion. In fact, this current waveform indicates the presence of high THD, in accordance with the experimental results. The SSCL method produces an approximately linear voltage drop due to resistive behavior of the superconductor tape. Therefore, the harmonic distortion of the current will be lower in this case than in the soft starter method.
101
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
We can view both the envelope and waveform of the SSCL current obtained by simulation in Figs. 17 and 18.
400
4.2. Torque analysis
300
Direct-on-line starting current (A)
1500
1232 A
1000
SSCL Torque (N.m)
200
100
0 1,0
0,8
0,6
0,4
0,2
0,0
Slip (pu) Fig. 21. SSCL electromagnetic torque.
400
Soft Starter Torque (N.m)
We can use the computational model to obtain the electromagnetic (EM) torque behavior during the starting of the IM using the soft starter and the SSCL methods. Electromagnetic torque is the torque that corresponds to the power transferred across the air gap. This is not the same as the output torque, or the load torque, that is available at the shaft because it does not account for friction, windage and stray load losses. It is well known that the EM torque (for any type of method) is quite oscillatory because the electromagnetic dynamics are faster than the mechanical dynamics. However, for the two starting methods presented in this paper, it is possible to compare one to the other. The current control mode of the soft starter has been set as a constant current mode during the startup of the IM. This was done to enable comparison of the two methods under similar conditions. In both methods, the current was limited to about half of the simulated DOL starting current (peak value). This can be observed in Fig. 19. Figs. 20–22 present the graphs of the DOL, SSCL and soft starter EM torque, respectively.
341 N.m
300 200 N.m
200
100
500 0
0
1,0
0,8
0,6
0,4
0,2
0,0
Slip (pu)
-500
Fig. 22. Soft starter electromagnetic torque.
-1000 -1500 -0,1 0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
Time (s)
The DOL maximum EM torque corresponds to approximately two times the SSCL maximum EM torque. The soft starter EM torque has plenty of noise from high frequency harmonic components of its current. For the soft starter, the high distortion of the current waveform results in an EM torque that is approximately 40% lower than the SSCL torque, for the same limited peak current. The models used in the simulations showed results that are consistent with our expectations. This is evidenced in the experimental tests and results described above.
Fig. 19. Simulated DOL starting current.
691 N.m
800
DOL Torque (N.m)
Some aspects should be highlighted for a proper comparison of the EM torque produced in each case.
600
5. Conclusions 400
200
0 1,0
0,8
0,6
0,4
Slip (pu) Fig. 20. DOL electromagnetic torque.
0,2
0,0
The performance of the new SSCL method confirmed the main goal of limiting the starting current of the IM. In experimental tests, a comparison with the DOL starting method showed a reduction of approximately 45%, which is close to the 50% value of the design specification. This slightly lower reduction indicates that the tape temperature was below 300 K, indicating safe operation of the superconductor module. THD measurement during the startup showed a maximum value of 36% for the soft starter, and a maximum value of 4.25% for the SSCL method. The harmonic pollution produced by the soft
102
F.B.B. Silva et al. / Physica C 507 (2014) 95–102
starter was neglected due to the short starting time. Altogether, in specific cases of industries where multiple IMs of large capacity are frequently started, and where the power systems have strict requirements for power quality indices, this high THD value of soft starters is a problem that must be addressed [13]. The EM torque produced in each method was investigated by simulation. The DOL maximum EM torque corresponds to about two times the SSCL maximum EM torque, and the soft starter method results in an EM torque that is approximately 40% lower than the SSCL torque. The models used in the simulations showed results that are consistent with our expectations. Based on simulation and experiments, we noted that the SSCL current is clearly less distorted than the soft starter current. In comparison, the superconductor starting current method offers a lower distortion of the current waveform, resulting in lower harmonic pollution during startup. In addition, we highlight other advantages of the SSCL such as greater robustness and a smaller number of electronic components. However, the SSCL method (using a YBCO 2G CC tape) does have some disadvantages. These include inhomogeneity of the superconducting tape, AC losses, losses by thermal conduction along the current leads, and high cost of operation with a cryogenic system required to maintain the superconductor at temperatures below its critical temperature [14,15]. The next step of this work is to test additional starting methods under different load conditions and under steady-state operation. Acknowledgements This work was supported by FAPES-PRONEX 48508675/2009 and CNPq-Universal 54694825/2011. The experimental results were obtained in the laboratory of electric motors in the Tubarão Complex/VALE SA with the help of the electrical engineer Flavio Helmer Ferro.
References [1] T. Hoshino, I. Muta, Load test on superconducting transformer and fault current limiting devices for electric power system, IEEE Trans. Magnetics 30 (4) (1994) 2018–2021. [2] D.U. Gubser, Superconducting motors and generators for naval applications, Physica C 392–396 (2) (2003) 1192–1195. [3] N. Amemiya, Q. Li, K. Takeuchi, T. Nakamura, M. Yagi, S. Mukoyama, Y. Aoki, N. Fujiwara, Effects of lateral-tailoring of coated conductor for AC loss reduction of superconducting power transmission cables, IEEE Trans. Appl. Supercond. 21 (3) (2011) 943–946. [4] M. Noe, M. e Steurer, High-temperature superconductor fault current limiters: concepts, applications and development status, Supercond. Sci. Technol. 20 (3) (2007) R15–R29. [5] Y. Hu, Z. Chen, Z. Chen, Y. Yuan, A novel method for the analysis of the impact of the high-power motor starting on grid, in: 4th International Conference on Electric Utility Deregulation and Restructuring and Power Technologies. IEEE, 2011, pp. 394–398. [6] C. Sen, Principles of Electric Machines and Power Electronics, third ed., John Wiley & Sons, New York, 2013. [7] L. Ghodhbani, B. Rebhi, F. Kourda, M. Elleuch, M. Abidi, Comparison among electronic start up methods for induction motors, in: IEEE 7th Int. Multi-Conf. on Syst. Signals and Devices, 2010, pp. 1–7. [8] A. Morandi, Fault current limiter: an enabler for increasing safety and power quality of distribution networks, IEEE Trans. Appl. Supercond. 23 (2013) 5604608. [9] C. Gandioli, P. Tixador, G.B. Mariani, Tests and simulations of different YBCO tape for FCL, IEEE Trans. Appl. Supercond. 22 (2012) 56003104. [10] W. Schmidt, H.-P. Kraemer, H.-W. Neumueller, U. Schoop, D. Verebelyi, A.P. Malozemoff, Investigation of YBCO coated conductors for fault current limiter applications, IEEE Trans. App. Supercond. 17 (2) (2007) 3471–3474. [11] C. Schacherer et al., Dissipated energy as a design parameter of coated conductor for their use in resistive fault current limiters, J. Phys: Conf. Ser. 97 (2008) 012193. [12] A. Riyaz, A. Iqbal, S. Moinoddin, S.M. Ahmed, H. Abu-Hub, Comparative performance analysis of thyristor and IGBT based induction motor soft starters, Int. J. Eng., Sci. Technol. 1 (1) (2009) 90–105. [13] G. Bhuvaneswari, S. Charles, M.G. Nair, Power quality studies on a soft-start for an induction motor, Transmission and Distribution Conference and Exposition, T&D. IEEE/PES, 2008, pp. 1–6. [14] S. Imparato et al., Experimental evaluation of AC losses of a DC restive SFCL prototype, IEEE Trans. Appl. Supercond. 20 (3) (2010) 1199–1202. [15] L. Ying, J. Sheng, B. Lin, L. Yao, J. Zhang, Z. Jin, Y. Li, Z. Hong, AC loss and contact resistance of resistive type fault current limiter using YBCO coated conductors, IEEE Trans. Appl. Supercond. 22 (2012) 5602404.