High Performance Parallel Single-Phase Converter Reconfiguration for Enhanced Availability Mohammad H. Hedayati Student Member, IEEE Indian Institute of Science (IISc) Bangalore 560012, India mh49929@gmail.com Vinod John Senior Member, IEEE Indian Institute of Science (IISc) Bangalore 560012, India vjohn@ee.iisc.ernet.in Abstract Paralleling power converters is a common practice in industries to enhance total power rating, reliability, and availability of the system. In case of fault occurring in systems with parallel converters, the faulty power converter can be isolated and the system can still be operated at reduced power level. In this paper, a grid-connected power converter consisting of two parallel H-bridge converter, with low ground leakage current, is considered. Two contingency configurations, that are also of low ground leakage current, are proposed to enhance the availability of the system. This is done by reconfiguring the power circuit to a single H-bridge, in the case of failure of the other bridge. The power converter is experimentally tested with the proposed configurations for experimental validation. The results show that, the second configuration has better performance in terms of power loss and current THD. Index Terms parallel converters, carrier interleaving, interphase inductor, LCL filter, grid-connected converters. I. INTRODUCTION Insulated gate bipolar transistor (IGBT) based gridconnected power converters are popular in application such as photovoltaic, battery chargers, renewables, wind power generation, and power factor corrections [1], [2]. Paralleling smaller power converters increases the system power level [3] [5]. It also increases the reliability of the system. This is achieved by ensuring higher system redundancy and availability [6]. Harmonic distortion injected to the grid can be reduced by interleaving the carrier [7] [10]. Interleaving the carrier of parallel converters shifts the harmonic components. This can results in cancellation of specific harmonics. Switching frequency loss reduction can be obtained by properly interleaving the carrier. The down side of the carrier interleaving is that, it can cause circulation of switching frequency current between the converters. This is due to turning on the top switch of one converter and bottom switch of the second converter, which belong to the same phase group, at the same time. The power converter also needs to operate with stringent requirements on ground leakage current, harmonic distortion, and EMI noise [11] [14]. In case of device failure, protection logic will act and shut down the whole system. However, in some cases it is possible to isolate the system from faulty devices and continue operation at reduced power levels. Different methods for reconfiguration of the power circuit are proposed in [15], Fig. 1. Grid-connected parallel single-phase power converters with interphase inductors and LCL filter for reduced ground leakage current. [16] so that the power converter can continue to operate even in case of a device failure. In [15], a matrix converter structure and a modulation technique for the remedial operation in case of opened switch faults and single-phase open circuits is proposed. This is done by reconfiguring the matrix converter topology with the help of a connecting device. A faulttolerant mechanism including topology and control strategy for multi-drive system based on indirect matrix converter is proposed in [16]. However, not much work has been reported on improving the availability of single-phase power converters with low ground leakage current characteristics. Fig. 1 shows a grid-connected power converter with high performance ground leakage current characteristics. The power converter consists of two H-bridge parallel connected converters. The H-bridge converters are connected together with the help of two inter-phase inductors, L int1 and L int2. LCL filter is used to interface the power converter with the power grid. In this paper, two contingency configurations are evaluated for grid-connected parallel single-phase power converter. The reconfiguration disconnects the faulty converter from the sys- 978-1-4763-7239-8/15/$31.00 ' 2015 IEEE 52
S1 S1 S3 S2 S3 S4 S2 (a) (c) (b) (d) Fig. 2. Contingency configurations. (a) The DPST switches introduced to reconfigure the circuit in Fig. 1 to Conf 1. (b) System configuration, Conf 1, when the second converter is faulty and isolated from the system. (c) The DPTP and SPST switches introduced to reconfigure the circuit in Fig. 1 to Conf 2. (d) System configuration, Conf 2, when the second converter is faulty and isolated from the system. tem and allows it to continue working at reduced power rating. In the first configuration, Conf 1, the inter-phase inductors are bypassed by the help of a double pole single throw (DPST) switches. The faulty power converter is isolated from the system by another DPST switches. In the second configuration, Conf 2, one of the the inter-phase inductor is used as a CM inductor. The other inter-phase inductor is disconnected from the system using double pole three position (DPTP) switches. The faulty power converter is isolated from the system by the help of a single pole single throw (SPST) switch. It is shown that in the first configuration the power converter is constrained in terms of employable modulation technique. The second configuration does not have this constraint, and unipolar or bipolar PWM can be used. A prototype converter system has been fabricated in the laboratory. The experimental results are provided which shows the effectiveness of the proposed method and the higher performance offered by Conf 2. II. POWER CONVERTER UNDER CONSIDERATION Fig. 1 shows a grid-connected parallel converter under consideration in this paper. In the shown topology, two H- bridge parallel connected converters are connected through inter-phase inductors, L int1 and L int2. LCL filter is used to filter the switching frequency component. The inter-phase inductor and the boost inductor of the LCL filter are integrated together as a single component, as suggested in [17], to reduce the size and the cost of the overall system. The LCL parameter are designed based on the guidelines provided in [18]. The link between the dc-bus capacitor midpoint and the LCL filter capacitor midpoint along with the capacitor C Mg to system are based on [19]. These are included into the system to reduce the ground leakage current, and to reduce the sensitivity of the power converters to nonidealities such as dead-time mismatch, interleaving angle error, delays in the gate signal circuit, and filter inductance mismatch. Unipolar PWM method with 180 interleaved carrier is used for the topology shown in Fig. 1. This results in low leakage current. In normal operation, the power is shared by the parallel converters. Typically, if a fault occurs in one of the H-bridge power converters, the whole system is shut down by the system protection logic. However, it is possible to isolate the faulty parts from the system and to keep the system running at reduced power level. To do so, switches are introduced into the power circuit to enable reconfiguration of the power circuit wiring. In this paper two contingency configuration are proposed in Section III. It is shown that Conf 2 leads to better performance in terms of, power losses and THD, compared to Conf 1. III. CONTINGENCY CONFIGURATIONS In this section the proposed contingency configurations and methods for changeover are studied for the case of failure of a single semiconductor switch of the inverter. It is shown that by 53
TABLE I THE STATES OF THE SWITCHES FOR RECONFIGURATION OF THE TWO TOPOLOGIES. Conf 1 Conf 2 Case S 1 S 2 S 3 S 1 S 2 S 3 S 4 (DPST) (DPST) (DPST) (SPST) (SPST) (DPTP) (DPTP) Normal Operation (Baseline) I I 0 I I I I Fault in Converter I 0 I I 0 I 0 II Fault in Converter II I 0 I I 0 II 0 DPST: I=Shorted, 0=Open, SPST: I=Shorted, 0=Open, DPTP: I=Position-I, I=Position-II, 0=Open. properly adding the switches into the power circuit, the power circuit wiring configuration can be changed, while retaining the low ground leakage current characteristics. These are used in the case of fault occurring in the semiconductor devices to isolate the faulty devices from the rest of the system. By doing so, the system can still be operated up to a reduced maximum power level of 50% of rated power. A. Contingency Configuration-1 (Conf 1 ) Fig. 2(a) shows the circuit diagram of the inter-phase inductors along with added switches to the power converter. The rest of the power circuit, which is not shown, remains unchanged as in Fig. 1. All three switches used in this configuration are DPST switches. Switches S 1 and S 2 are added, to the power circuit, to isolate the faulty H-bridge power converter. In this case, when only one of the converter is operating, the interphase inductors offer a high impedance path for the output current. This results in high voltage drop across the interphase inductors. Hence, it is necessary to bypass the interphase inductors. To do so, S 3 is added to the power circuit. In this case, the mutual fluxes in the core cancel each other and the inter-phase inductors act as a short circuit. Fig. 2(b) shows the case when the second H-bridge power converter is faulty and is isolated from the system. This circuit configuration after the changeover is as suggested in [19], which results in low ground leakage current and EMI when excited using bipolar PWM. The ripple current frequency, in the case of bipolar PWM, is the same as the carrier frequency. Hence, the power losses in the filters are expected to be more with respect to the normal operation. Unipolar PWM technique cannot be used with this configuration. This is due to high circulating current that will flow within the power converter [19]. Bipolar PWM method is the suitable modulation technique for this configuration. The states of the switches to reconfigure for operation as Conf 1, in case of fault in converter I, or fault in converter II, are tabulated in TABLE I. B. Contingency Configuration-2 (Conf 2 ) In this configuration, one of the the inter-phase inductor is reconfigured to act as a common-mode (CM) inductor in the power circuit. The other inter-phase inductor is isolated by the help of either S 3 or S 4. To do so, it is necessary to open the inter-phase inductor midpoint. S 3 and S 4 ensure that the inter-phase windings are connected such that it offers high impedance to the CM current and very small impedance to the output current. These switches are DPTP and have three positions I, II, and open (0). S 1 and S 2 are SPST switches. These are added to the power circuit to isolate the faulty H- bridge from the rest of the system. The states of the switches to reconfigure for operation as Conf 2, in case of fault in converter I, or fault in converter II, are tabulated in TABLE I. Fig. 2(c) shows the circuit diagram along with added switches to reconfigure the power converter to Conf 2. The rest of the power circuit, which is not shown, remains unchanged as in Fig. 1. Fig. 2(d) shows the case when the second H-bridge converter is faulty and isolated from the rest of the system. This circuit configuration after the switch changeover is as suggested in [20]. Unipolar as well as bipolar PWM method can be used in this configuration. Unipolar PWM method is beneficial due to reduction of the current ripple and current THD. This also reduces the power loss compared to Conf 1. The OM connection along with the capacitor C Mg keep the ground leakage current low. The circulating current, due to usage of unipolar PWM along with OM connection, is limited by the help of inter-phase inductor. IV. EXPERIMENTAL RESULTS A 7.5kVA parallel single-phase power converter, as shown in Fig. 1, is fabricated in the laboratory. The photograph of the power converter, showing labelled parts, is in Fig. 3(a). Fig. 3(b) shows the integrated inter-phase inductor with the boost inductor, along with the switches used for changing the converter configuration. An ALTERA Cyclone II fieldprogrammable gate array (FPGA) is used for control. The setup is tested at full rating as a static compensator (STAT- COM). The grid voltage and current are shown in Fig. 4. The total harmonic distortion (THD) of the grid voltage and current are measured to be 2.19% and 1.36% respectively. Three cases are considered for experimental evaluation. Case A, Baseline configuration, normal parallel operation with unipolar PWM and 180 interleaving angle, configuration shown in Fig. 1. Case B, Conf 1, one of the power converter is operating with bipolar PWM method, the configuration shown in Fig. 2(b). 54
Current Voltage C1 C2 C1 DC1M 100V/div 0.0V offset C2 DC1M 20.0A/div 0.0V offset Timebase 500 us 5.0ms/div 100kS 2.0MS/s Fig. 4. Grid voltage and current of the power converter at 7.5kVA while operating as a STATCOM. 100V/div, 20A/div, and 5ms/div. (a) (b) Fig. 3. Experimental prototype. (a) The 7.5kVA power converter. (b) Integrated inter-phase inductor with the boost inductor, along with the switches used for configuration changeover. Case C, Conf 2, one of the power converter is operating with unipolar PWM method, the configuration shown in Fig. 2(d). A. Leakage Current The leakage current is the current which is injected to ground. It is sensed by measuring the difference between phase current and neutral current. The setup is tested with the aforementioned cases A, B, and C and the leakage current is measured. The tests have been carried out at full load, (7.5kVA), for case A, and at half load, 3.75kVA, for cases B and C. The results are shown in Fig. 5. Fig. 5(a) shows the spectrum of the leakage current for case A. The rms value of the leakage is measured to be 12.6mA. Fig. 5(b) and Fig. 5(c) show the leakage current spectra for cases B and C respectively. The rms values of the leakage currents are 15.5mA and 12.8mA respectively. Standard VDE V 0126-1-1 [12] specifies the maximum leakage current of power converter to be 300mA. If the power converter leakage current is higher than that, it should be disconnected within 0.3s. It can be seen, from the experimental results, that all the three cases have similar level of leakage current and meet the VDE standard recommendation. B. Power Loss The power converter is tested, at different loading condition, and the power losses are measured. In case A, the load is varied from no load to 7.5kVA. In cases B, and C, the load is varied from no load to 3.75kVA. Power loss is measured using a YOKOGAWA WT1600 digital power meter. The results are shown in Fig. 6. It can be seen that the losses in case B are higher in comparison with case A and C. It was expected as the bipolar PWM is used in this case. The losses in cases A, and C are close. However, the losses in case C is lesser up to 3kVA, and for higher than 3kVA, case A has lesser power loss. In case C, Conf 2, one of the integrated inductor is disconnected. Hence, the power losses in the magnetic component is saved and the power losses are reduced at the lower power level. This test suggests that, if automated switches are used in the setup for changing the configuration, at light load it is beneficial to reconfigure the setup to Conf 2. This helps in increasing the efficiency. C. Current THD The power converter is tested, in the mentioned cases in Section IV, with different loading condition. The results are shown in Fig. 7. It can be seen that, all the cases have high THD at light load. This is due to the presence of lower order harmonics in the grid voltage and light loaded conditions. The light loading results in low fundamental current. Hence, it causes the current THD to be high. As the loading increases, the THD performance is improved. It can be seen that, the THD values in the baseline and Conf 2 cases are the same. The reasons are, these two cases have similar filtering and the same effective switching frequency. Due to usage of unipolar PWM method, the effective switching frequency is double the switching frequency. In Conf 1, the effective switching frequency is same as switching frequency, due to usage of bipolar PWM method. Hence, for the same loading condition, 55
600 500 Power loss (W) 400 300 200 FFT(C1) 20dB/div 25kHz/div (a) Timebase 0.0 ms 2.0ms/div 10kS 500kS/s 100 0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Reactive Power (kva) Fig. 6. Power converter power loss versus converter loading for cases A, B, and C. 40 35 30 FFT(C1) 20dB/div 25kHz/div (b) Timebase 0.0 ms 2.0ms/div 10kS 500kS/s THD(%) 25 20 15 10 5 0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Reactive Power (kva) Fig. 7. Current THD of power converter versus loading for cases A, B, and C. FFT(C1) 20dB/div 25kHz/div (c) Timebase 0.0 ms 2.0ms/div 10kS 500kS/s Fig. 5. Ground leakage current spectra. (a) Case A, configuration shown in Fig. 1 at 7.5kVA loading. (b) Case B, configuration shown in Fig. 2(b) at 3.75kVA loading. (c) Case C, configuration shown in Fig. 2(d) at 3.75kVA loading. 20dB/div, and 25kHz/div. the THD values in case Conf 1 are higher in comparison with Conf 2 and baseline. V. CONCLUSION Two contingency configurations are evaluated for the case of fault occurring in the semiconductor devices of a parallel single-phase grid-connected converter. Manual changeover switches are added to the power converter to reconfigure the power circuit. The experimental tests are carried out on a 7.5kVA laboratory prototype power converter. It is shown that, in case of failure of one of the H-bridge, the system can work at reduced power level. This enhances the availability of the system. The leakage current for baseline, Conf 1, and Conf 2 are measured to be 12.6mA, 15.5mA, and 12.8mA respectively. All the cases have low leakage current and meet the specification of the VDE [12] standard. Changing the configuration to case B, Conf 1, results in higher power loss and higher current THD. Reconfiguration of the power circuit to case C, Conf 2, results in lower power loss in comparison to baseline. The current THD are the same for the cases Conf 2 and baseline. The second configuration shows better performance with respect to current THD and power loss. VI. ACKNOWLEDGEMENT We would like to thank the Department of Heavy Industry (DHI) Government of India for funding this project and supporting us throughout the work. REFERENCES [1] S. Bernet, Recent developments of high power converters for industry and traction applications, IEEE Trans. Power Electron., vol. 15, no. 6, pp. 1102 1117, Nov 2000. 56
[2] J. Rodriguez, J. Dixon, J. Espinoza, J. Pontt, and P. Lezana, PWM regenerative rectifiers: state of the art, IEEE Trans. Ind. Electron., vol. 52, no. 1, pp. 5 22, Feb. 2005. [3] S. J. Chiang, C. Y. Yen, and K. Chang, A multimodule parallelable series-connected PWM voltage regulator, IEEE Trans. Ind. Electron., vol. 48, no. 3, pp. 506 516, Jun 2001. [4] T. Kawabata and S. Higashino, Parallel operation of voltage source inverters, IEEE Trans. Ind. Appl., vol. 24, no. 2, pp. 281 287, Mar 1988. [5] M. Baumann and J. Kolar, Parallel connection of two three-phase three-switch buck-type unity-power-factor rectifier systems with dc-link current balancing, IEEE Trans. Ind. Electron., vol. 54, no. 6, pp. 3042 3053, Dec 2007. [6], Parallel connection of two three-phase three-switch buck-type unity-power-factor rectifier systems with dc-link current balancing, IEEE Trans. Ind. Electron., vol. 54, no. 6, pp. 3042 3053, Dec 2007. [7] X. Mao, A. Jain, and R. Ayyanar, Hybrid interleaved space vector pwm for ripple reduction in modular converters, IEEE Trans. Power Electron., vol. 26, no. 7, pp. 1954 1967, July 2011. [8] L. Asiminoaei, E. Aeloiza, P. Enjeti, and F. Blaabjerg, Shunt activepower-filter topology based on parallel interleaved inverters, IEEE Trans. Ind. Electron., vol. 55, no. 3, pp. 1175 1189, March 2008. [9] T. P. Chen, Zero sequence circulating current reduction method for parallel HEPWM inverters between ac and dc bus, IEEE Trans. Ind. Electron., vol. 59, no. 3, pp. 290 300, Jan. 2012. [10] J. Prasad and G. Narayanan, Minimization of grid current distortion in parallel-connected converters through carrier interleaving, IEEE Trans. Ind. Electron., vol. 61, no. 1, pp. 76 91, Jan 2014. [11] IEEE recommended practices and requirements for harmonic control in electrical power systems, IEEE Std 519-1992, 1993. [12] Automatic Disconnection Device between a Generator and the Public Low-Voltage Grid, DIN Electrotechnical Standard DIN VDE 0126-1-1, 2005. [13] Electromagnetic compatibility (EMC) - part 4-6: Testing and measurement techniques - immunity to conducted disturbances, induced by radio-frequency fields, IEC 61000 Edition 3.0, 2008-10. [14] Industrial, scientific and medical radio-frequency equipment electromagnetic disturbance characteristics limits and methods of measurement, CISPR11 Edition 5.0, 2009-05. [15] S. Kwak and H. Toliyat, An approach to fault-tolerant three-phase matrix converter drives, IEEE Tran. Energy Convers., vol. 22, no. 4, pp. 855 863, Dec 2007. [16] Y. Mei and Z. Li, A fault-tolerant mechanism for multi-drive system based on indirect matrix converter, in IECON 2011-37th Annual Conference on IEEE Industrial Electronics Society, Nov 2011, pp. 1036 1040. [17] M. Hedayati and V. John, A novel integrated cm inductor for parallel converters with interleaved carrier, in Power Electronics, Drives and Energy Systems (PEDES), 2014 IEEE International Conference on, Dec 2014, pp. 4 6. [18] P. Channegowda and V. John, Filter optimization for grid interactive voltage source inverters, IEEE Trans. Ind. Electron., vol. 57, no. 12, pp. 4106 4114, Dec. 2010. [19] M. Hedayati and V. John, Filter configuration and pwm method for single phase inverters with reduced conducted emi noise, IEEE Tran. Ind. Appl., vol. PP, no. 99, pp. 1 1, 2015. [20] D. Dong, F. Luo, D. Boroyevich, and P. Mattavelli, Leakage current reduction in a single-phase bidirectional ac-dc full-bridge inverter, Power Electronics, IEEE Transactions on, vol. 27, no. 10, pp. 4281 4291, Oct 2012. 57