Thank you for visiting Nature.com. The browser version you are using has limited CSS support. For best results, we recommend using a newer browser (or turning off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we will display the site without styles and JavaScript.
Recently, many high gain topologies have been developed. However, there is a need for a converter with high gain and fault-tolerant performance. This paper proposes a fault-tolerant reconfigurable secondary boost converter for DC microgrids. In this new topology, level 2 redundancy is achieved by eliminating switch and capacitor faults. The performance of the converter under normal operating conditions and in the reconfigured mode is discussed. The obtained topology can provide the same voltage gain even in the reconstruction mode. The proposed topology shows the best performance in the reconstructed state. In this state, the voltage across the input and output capacitors decreases. Using the reliability handbook, the reliability analysis of the capacitors in both states was conducted and compared. Finally, the performance of the topology in normal and reconstructed states was verified by constructing a 1 kW hardware setup. The results show that the secondary boost converter in the reconfigured state operates without changing the gain of the converter and the voltage across the capacitor decreases.
In 2023, solar power will grow faster than all other power generation technologies. Last year, solar PV accounted for two-thirds of new renewable energy generation, with a power generation growth rate of 24%, the highest among all power generation technologies. Solar power plays an important role in the DC microgrid architecture. By using multiple power converters in the DC microgrid structure, they are more responsive to changes in transient conditions1. For the flexible operation of the DC microgrid, a multi-port DC-DC converter is proposed to control the DC link voltage2. Robust control schemes in DC microgrids are critical for deployment in rugged terrain3. A robust fault-tolerant series resonant converter is designed for DC microgrids and is designed to overcome the faults of semiconductor switches4. This deliberately demonstrates the need for smooth control and robustness in DC microgrid structures, especially in power conditioning devices. The researchers’ goal is to pave the way for the development of reliable energy harvesting interfaces for solar panels and other renewable energy sources5.
Recently, many topologies have been developed and presented in the literature 1, 2, 6, 7 for renewable energy applications. On the other hand, little attention has been paid to the research on the fault-tolerant structures of power converters. The service life of photovoltaic/wind turbines is about 20 years, while the reliable service life of power converters available in the market is relatively short. Therefore, the power interface unit needs to be replaced twice during the reliable operation of the hybrid renewable energy system. Therefore, hybrid renewable energy systems require the most reliable topology. The demand for power converters for renewable energy generation has increased rapidly in the past few years8.
In the design stage of the power converter, fault tolerance techniques should be incorporated into the most critical applications9. Regarding the failure analysis in the literature, it was concluded that after the capacitors, the second most prone to failure components are printed circuit boards (26%) and switches (21%)10. By adding redundant components to the converter11, 12, 13, the reliability of the converter can be increased. The failure of active semiconductor switches will lead to catastrophic and irreversible failure of the entire system14. The main component of the power converter that fails is the capacitor due to the charging cycle and degradation of the dielectric material. The fault-tolerant structures are reported in the literature to target switch failures rather than capacitor failures11,15,16. In all these topologies, switch faults are diagnosed11, 15, 16 and measures to correct them are proposed. However, capacitor failures (especially output ones) are not monitored and taken into account because the capacitor voltage is basically equal to the converter output voltage.
In this proposal, a suitable topology will be proposed and presented to meet the requirements of an efficient DC microgrid. In the literature, hardware redundancy is mainly proposed for fault handling and has been proven to be an effective method to improve the reliability of the converter. In 17, a fault-tolerant and reliable structure was achieved by cascading multiple quasi-Z converter modules. When a converter switch fails, the faulty module is isolated from the cascaded converter and continues to operate using a relay. This fault-tolerant architecture has multiple redundant components, which increases the initial investment and cooling cost of the converter.
In 18, a reconfigurable step-up/step-down converter with 1-degree redundancy, i.e., diagnosing and correcting switching faults, is proposed. In this topology, when a switch fails, the step-up configuration is reconfigured into a step-up-step-down configuration. However, fault-tolerant topologies with reconstruction methods have not been fully explored. According to the current research in the proposed field, reconfiguration topologies with 2-degree redundancy (switches and capacitors) have not been studied in the literature, which will serve as the purpose of creating a new topology with 2-degree redundancy and reconfiguration methods. . In addition, a power management strategy will be proposed to provide an interactive working platform for renewable energy sources and coordinate their operation according to the requirements of the microgrid6,7. Power management is indispensable in microgrids since it ensures reliable operation even under failures and unexpected situations19,20. This study is currently valued as it ensures smooth and uninterrupted operation of systems by prioritizing sources.
In India, 2.4% of households still do not have access to electricity. These rural homes will benefit from renewable DC systems. Villages located in rugged terrains such as deserts, islands and mountains require reliable power supply systems as replacing faulty parts is difficult and time-consuming. Therefore, it is recommended to design a robust system in which the design of the power stabilization device is critical as it is more susceptible to failure. If the reliability analysis of power converters is adequately studied, suitable fault tolerance techniques can be used to control faults in power components21.
Table 1 provides a literature review of the converter types considered for fault tolerance studies and their fault recovery components. Research on integrating refactoring capabilities with fault-tolerant operations is also reviewed and presented. It is evident from this study that most of the switch faults have been analyzed. In the literature, fault diagnosis and identification have been studied more than the creation of fault-tolerant reconfigurable topologies. It is also noted that there is significant scope for research on proposing reconfigurable fault-tolerant topologies without changing the key characteristics of the converter.
In the literature, the fault tolerance of single-degree redundancy is mainly discussed and tested. According to the current research background in the proposed field, reconstructed topologies with 2-degree redundancy have not been investigated in the literature, which would be the main goal. The innovation in the proposed configuration is the creation of a high-gain topology and the development of a fault-tolerant structure that allows reorganization in the event of a fault without changing the voltage gain. An additional feature of the topology is the reduction of input current ripple, which is necessary for systems operating on renewable energy sources22.
The main contributions of this study are summarized as follows: 1. A novel reconfigurable fault-tolerant topology with distinctive features is proposed. 2. The proposed topology is evaluated for reliability and compared with a conventional secondary boost converter. 3. The proposed RFTQB topology is compared with existing topologies in the literature to highlight its salient features. 4. To test the proposed topology under normal operation and reconstruction modes on a 1 kW prototype. 5. The experimental setup is verified by conducting experiments using a programmable DC power supply and solar PV modules.
The paper is organized as follows: Section 2 presents the schematic of the reconstructed secondary boost converter with fault tolerance. Section 3 discusses the steady-state analysis of the reconstructed topology in normal and restored states and presents the corresponding waveforms and diagrams. In the same section, the efficiency, dynamics and sensitivity analysis under non-ideal voltage gain are performed. The proposed topology is compared with a conventional secondary boost converter and its superiority is highlighted. Section 4 investigates and compares the reliability of the proposed converter in normal and reconstructed states. Section 5 presents a comparative analysis with similar fault-tolerant topologies reported in the literature. Simulation and hardware results are obtained to validate the theoretical study of the proposed topology and the results are described in Sections 6 and 7, respectively. Finally, the results are summarized in Section 8. The terminology for this study is presented in the supplementary file.
Figure 1 shows the average price per watt of solar panels and global solar PV module production over time (2010-2020). It shows that solar module prices have fallen by up to 90%, while solar cell production has grown by up to 400%. Researchers are working to lay the foundation for developing robust energy harvesting interfaces for solar panels and other renewable energy sources. Figure 1b shows the failure rate of power converter components. As can be seen from this figure, capacitors and switches account for more than 50% of power converter component failures. Given this, it is recommended to use 2 degrees of redundancy for a secondary boost converter with additional functionality.
Figure 2 provides an overview of the proposed fault-tolerant architecture configuration. Figure 2a shows the proposed structure obtained by modifying a conventional secondary boost converter. In the proposed configuration, the inductors (L1 and L2) are moved to the bottom bus of the circuit. Similarly, the diode D1 on the input side is moved between the emitter of the IGBT and the inductor L1. The diode D2 is moved between the negative terminal of the input capacitor C and the inductor L1. Finally, the output diode is moved between the negative terminal of the output capacitor Co and the inductor L2. By moving these components, a new fault-tolerant converter with level 2 redundancy is designed. This topology is more suitable for stand-alone applications.
(a) Derivation of the proposed structure (b) Reconfigurable fault-tolerant quadratic boost converter (RFTQB) (c) Consideration and analysis of RFTQB converter with faulty components (d) RFTQB converter in normal state (e) RFTQB converter in reconfigured state.
Figure 2b shows the schematic diagram of the reconfigured fault-tolerant quadratic boost converter (RFTQB). Figure 2c shows the faulty components of the proposed converter. The schematic diagram of the normal state and reconfiguration state of the RFTQB converter is shown in Fig. 2d, e.
Figure 2d shows a conventional secondary boost converter with the components connected to the bottom rail. The RFTQB topology is based on the traditional secondary boost converter. The diodes (D1, D2, and Do) and the inductor (L1 and L2) are moved to the bottom rail as shown in Figure 2e. Additional switches SWR, input capacitor CR, and output capacitor CoR are added to allow the topology to operate in a reconfigured state. As a result, the circuit configuration is designed to convert to a floating output signal. This type of circuit is more suitable for applications such as off-line applications with a battery. Three fuses have been added in series with the switch and capacitor to isolate the component from fault occurrence. Switch SWR is added in parallel with the main switch SW and becomes active only when the main switch fails. In literature 12, switch failures are eliminated by using additional switches to provide circuit redundancy.
Compared with the converter proposed in Article 17, the proposed topology uses redundant components instead of redundant circuits/modules to eliminate the failure of a single component. 17, sensors and relays are used to design fault-tolerant structures. The output voltage of the cascaded quasi-Z-source converter 17 is monitored by sensors. When the voltage decreases, the relay isolates the faulty module and increases the duty cycle of the healthy converter.
In the proposed topology, three more fault-tolerant components (2 capacitors and 1 switch) are added as a backup to make the converter fault-tolerant. In addition to fault-tolerance, the proposed topology also has other reconfigurable features such as undervoltage reduction on the input and output capacitors. This is verified through steady-state analysis in the next section.
In this mode, the topology operates in the same way as a traditional secondary boost converter. The voltage across the diode, switch, and capacitor is similar to that of the secondary boost converter.
This mode is active only when switches and capacitors fail. Figure 3 shows how the topology behaves when each component fails. The redundant components in the circuit are the conductive input capacitor (CR), the output capacitor (CoR), and the switch (SWR). After a fault occurs, redundant components appear. The three faults and their circuit configurations are shown in Figures 3a-c.
It is observed that the voltage across the switch is equal to the output voltage in both modes of operation. By adding a voltage multiplier (diode-capacitor) after the switch and before the load, the voltage across the switch can be removed. This will increase the topological gain. This configuration is mainly aimed at reducing the load on the capacitor. This voltage reduction will be verified in the next section by performing a steady state analysis.
To check the performance of the RFTQB converter, a steady-state analysis is performed using the volt-second equilibrium law for the normal mode and the restored mode. In both modes, the voltages on the inductor (L1 and L2) are observed, which are given in Table 2. The operating modes of the current flow in the circuits are presented. After applying the volt-second equilibrium principle, a general expression is obtained.
Applying the second voltage balance principle, the on and off cycles are calculated for normal operating conditions.
Finally, the voltage conversion ratio of the proposed topology in normal operation is calculated as
Similarly, the principle of the second voltage balancing is applied to the on-off cycle reconfiguration mode of operation, the voltage on the backup input capacitor is
The constant voltage gain of the proposed reconfigurable fault-tolerant secondary step-up converter in the reconfiguration mode is calculated, similar to (2).
It can be seen that the voltage gain of the converter remains constant in all operating modes in which the fault occurs. It can also be seen from Figures 2a-c that the voltage gain of the converter is the same as (2) in case of input, output and switch faults. In (2), Vdc is the solar panel voltage. Similarly, the voltage across the switch and diodes (D1, D2 and Do) remains constant as shown in Table 2. Figure 4 shows the voltage across the diode and capacitor in normal and reconfiguration modes. The voltage across the capacitor in the configuration changes when a fault occurs.
(a) Voltage across the diode and capacitor in normal state (b) Voltage across the diode and capacitor in reconfiguration state (c) Switching circuit of the proposed fault-tolerant converter in normal mode (d) Proposed fault tolerance in reconfiguration mode Switching circuit of the proposed fault-tolerant converter in normal mode converter (e) Root plot (e) Bode plot (g) Change in voltage gain for different values of duty cycle and parasitic resistance (h) Losses of power components.
Among them, GV is the voltage gain of the converter, equal to 1/(1−D)2. RL and RC are the internal resistances of the passive components.
Consider Po = 200 W, Vdc = 50 V, Vo = 200 V, Ro = 200 Ω, RF = 0.083 Ω, VF = 2.5 V, RCE = 10 mΩ, RL = 160 mΩ, RC = 290 mΩ, and COES = 0.4 nF. Calculate the efficiency of the fail-safe power converter. Its theoretical efficiency is 90%. The loss distribution is shown in Figure 4g.
The dynamic study of the proposed converter is performed using the switching flow graph (SFG) method. Fig. 4c, d show the SFG of the fault-tolerant converter in the normal mode and reconstruction mode, respectively.
Figure 4e shows the root locus of the proposed converter. The figure shows four trajectories with four poles. All poles lie on the imaginary axis. Two trajectories move to the left of the plane, and the other two trajectories move to the right of the plane. Analysis of the graph shows that the converter is in a minimally stable state.
Figure 4f shows the Bode response of the proposed converter. It is a type 0 system with four poles. The gain reaches 0 dB and the phase reaches 180 degrees at different frequencies. Since the system is type 0, the amplitude diagram remains constant at low frequencies and the phase at these frequencies tends to 0 degrees.
To study the change in voltage gain with duty cycle, a sensitivity analysis was performed. Assuming that the two inductors have the same internal resistance, study the change in voltage gain when the internal resistance of the inductor is different. Equation (30) is simplified if only the parasitic resistance of the inductor is taken into account.
Figure 4z shows the change in non-ideal voltage gain for different values of the parasitic resistance of the choke and the duty cycle. As can be seen from this figure, the higher the duty cycle, the more obvious the change in voltage gain.
This section analyzes the reliability of the proposed topology, focusing on the failure rates of the input and output capacitors. Failure Rates of Fixed Capacitors, Aluminum Capacitors, and Electrolytic Capacitors
where T is the ambient temperature in °C and s is the ratio of the capacitor’s working voltage to its rated voltage. The basic failure rate is often related to the effects of temperature and electrical stress on components.
The voltage ratios of the input and output capacitors of the secondary boost converter in the normal operation mode of the proposed converter are (1) and (2), respectively. Similarly, the voltage ratios of the input and output capacitances of the proposed converter in the restored operation mode are (3) and (4), respectively. The comparison of the capacitor voltage ratio is shown in Figures 5a, b. Figure 5a shows the comparison of the input capacitor voltage ratio in the normal and restored modes of the proposed topology. Similarly, Figure 5b shows the comparative study of the output capacitor voltage ratio in the normal secondary boost and reconfiguration modes. Using expression (8), the base failure rate λb can be determined at different temperatures and duty cycles. The comparison of the results is shown in Figures 5c, d. Figure 5c shows the base failure rate of the input capacitor for 0.2 < D < 0.8 and different temperatures. Figure 5d shows the base failure rate of the output capacitor for 0.2 < D < 0.8 and different temperatures. From this comparative study, it can be seen that the reconfigurable mode of operation has a lower base failure rate of both the input capacitor CR and the output capacitor CoR.
Reliability Assessment (a) Input Capacitor Voltage Rating (b) Output Capacitor Voltage Rating (c) Input Capacitor Baseline Failure Rate (d) Output Capacitor Baseline Failure Rate (e) Selected Capacitor Rating Baseline Failure Rate Comparison (f) Failure Rate Comparison of Capacitors with Selected Ratings (g) Switch Failure Rate Analysis (h) Diode Failure Rate Analysis (i) Inductor Failure Rate Analysis (j) Capacitor Failure Rate Analysis.
The failure rate of a capacitor can be determined using the expression (13). In this expression, πCV is the capacitance factor, πE is the environmental factor, and πQ is the quality factor. πCV is obtained using the capacitance value of the capacitor. It is calculated using the following formula
where C is the capacitance of the capacitor. Since the ground-mounted (GM) case is considered in this study, the environmental factor πE is chosen to be 12 for GM from the guideline. The quality factor πQ is considered to be 0.03 of the quality S. For this analysis, the 1 kW converter has Vdc = 50 V, Vo = 200 V, T = 25 °C, and D = 0.5. The base failure rate and the failure rate of the capacitor were calculated and compared (Fig. 5e, e). Figure 5(e) is obtained by estimating the base failure rate of the capacitor according to the expression shown in Figure 5j. Finally, the failure rate of the capacitor shown in Figure 5f is determined using the expression shown in Figure 5j. From this comparison, it is clear that the input and output capacitors in the reconfigured state have a lower failure rate compared with the conventional secondary boost converter.
Figure 5g-j shows the failure rate analysis of the power converter components. Figure 5g-j shows the expression of the failure rate λ for each component. The failure rates of switches, diodes, inductors and capacitors are 36.23 FIT, 2.19 FIT, 0.0276 FIT and 0.00244 FIT, respectively. The overall failure rate of the converter is 38.45 FIT. The mean time to failure is calculated by taking the reciprocal of the overall failure rate.
Perform a cost analysis of the converter considering reliability. The total cost of the power converter includes the initial installation cost, downtime, and power loss. The cost of the converter system is included in the cost of installing the converter. According to 23, the downtime cost is assumed to be $2/hour considering the financial loss. The expression for the downtime cost is obtained by assuming that the time required to replace the failed equipment is one week. However, in this topology, the downtime cost is negligible due to the availability of redundant equipment.
Furthermore, let us assume that the cost of power losses is 10 cents/kWh. Given this, the total cost of power losses is given by
The switch, diode, inductor, and capacitor cost $120, $3, $15, and $10, respectively. The installation cost of the converter without redundant components and with redundant components is $194 and $344, respectively. The total cost of the power converter in normal and reconfiguration modes is $281, since the number of components operating in both modes remains the same. Compared to topology 23, this fault-tolerant topology incurs downtime costs, since the converter is reconfigured without interrupting operation during the failure phase.
In this section, we compare the proposed RFTQB topology with the converter proposed in 18. The RFTQB topology eliminates switch and capacitor faults, while the converter obtained in 18 is suitable for switch faults. It has been noted in the literature that capacitors in power converters are more prone to failure. The RFTQB topology takes into account the input and output capacitor faults as well as switch faults. In addition, the reconstructed topology proposed in 18 is converted from buck and boost converters to buck-boost converters, while the RFTQB topology is restored to the same topology without breaking the topology due to switch and capacitor voltage faults and current gain.
The proposed RFTQB topology is compared with the FT topology proposed in 18 and 24, and its illustration is shown in (Fig. 6a-c). It is worth noting in Figure 6a that the obtained FT topology with single-level redundancy changes the topology configuration. Similarly, the fault-tolerant topology proposed in 24 is taken into account for comparison, and its reconfiguration during switch failure is shown in Figure 6b. In this topology, a buck-boost converter is integrated with a positive-output Luo converter to eliminate switching errors. This results in a high-gain DC-DC converter with two switches. In case of switch S1 failure, the proposed converter is reconfigured as a positive-output Luo converter, and in case of switch S2, it is reconfigured as a boost converter. This topology also eliminates only switch failures.
For the RFTQB topology, it is important to note that the reconfigured topology does not change the topological configuration and voltage gain of the converter when switches and capacitors fail (two-level redundancy), as shown in Figure 6c.
The RFTQB converter was simulated in MATLAB/Simulink and the simulation results are discussed in this section. Figure 7(a.e.) shows the simulation results of the proposed topology. For this simulation, we consider Vdc = 50 V, Vo = 200 V, fs = 20 kHz and D = 0.5. Figure 7a shows that the output voltage remains constant in both operating modes. Figure 7b confirms the capacitor voltage reduction in the reconfiguration mode by showing the voltages on the input and output capacitors.
Post time: Nov-19-2024