SiC vs. Si – Potentials of Wide-bandgap Devices for Performance Improvement in Energy Conversion Systems

Get Complete Project Material File(s) Now! »

Power Electronics for Photovoltaics

Although being the most known element in photovoltaics, the PV generator (or PV cell) is not the only fundamental component of a PV system. In reality, the voltage and current produced by a PV generator are DC quantities. Thus, to be able to inject power into the grid or feed an AC load, a DC-AC converter (or inverter) is often employed. Furthermore, due to its non-linear i–v characteristic4, the PV cell presents a single maximum power point (MPP), which must be constantly tracked by the inverter to maximize energy yield during operation. In the literature, numerous maximum power point tracking (MPPT) techniques can be found [5], [6]. Since the PV voltage varies during the day according to the weather and cell’s temperature conditions, the inverter is also responsible to dynamically ensure a constant RMS AC output voltage. PV inverters can be designed in single- and three-phase configurations. Due to the presence of a low-frequency (100Hz) fluctuating power component in single-phase systems, this inverter category is usually limited to a few kW and plays a minor role in the overall PV inverter market share [4] (mainly due to cost reasons). On the other hand, three-phase inverters account for more than 95% of the market share [4] and are the preferred solution in commercial and utility sectors [3].
When operating in grid-tied mode, the inverter must be able to: i) insure the DC-AC conversion with minimum power quality requirements [7]; ii) offer grid support injecting or absorbing reactive power, complying with grid codes and regulations [8]; iii) disconnect itself from the grid, when necessary (anti-islanding)
In both architectures, a low-voltage/medium-voltage (LV/MV) step-up transformer allows the PV power plant to be connected to the local MV grid.
Concerning the lifetime aspect of PV systems, there is a considerable gap between the typical lifespan of inverters (5 to 15 years [3], [12]) and that of PV panels (around 25 years [13]). From an operation and maintenance point of view, inverters are considered to be the weakest components in PV systems. In [14], it is reported that inverters were responsible for 37% of unscheduled maintenance events in a 50MWp PV plant, accounting for 59% of the associated overall repair costs. In [15], similar results are presented about the high cost of inverter failures, which comprises not only the service cost itself, but also the energy losses due to the PV power plant downtime. When identifying the root causes of inverter reliability issues, power semiconductors and DC-link capacitors (often electrolytic type) represent the most common failure sources [16]. Taking all this into consideration, the improvement of PV inverter reliability remains a topic to be addressed in order to reduce operation and maintenance costs, as well as energy losses due to inverter downtime.

Research Motivations and Objectives

In Figure 1.3 is depicted a chart with three possible actions to be taken at the inverter design level to increase the energy yield and reduce costs of utility-scale PV power plants.
Another possible action to facilitate capping costs in PV domain is the implementation of wide-bandgap (WBG) semiconductors, especially Silicon Carbide (SiC) and Gallium Nitride (GaN) devices. Due to their superior physical and electrical properties, the efficiency of energy conversion systems can be sensibly improved, lowering PV LCOE. Thanks to their aptitude to operate at superior switching frequencies, the inverter power density can be increased, helping to reduce the PV plant mounting and shipment costs, for example. Furthermore, WBG devices present outstanding on-state characteristics, allowing the power electronic engineer to redesign and simplify standard converter topologies. For example, by reducing the number of active switches and the associated complexity, while improving reliability and possibly lifetime. However, the major drawbacks of such semiconductors are related to their early stage of technological maturity, leading to: i) relatively high cost, which might be overcome to some extent by massive-scale production and ii) reliability issues. Concerning SiC devices, the reliability-critical aspects have been reported to involve gate-oxide long-term reliability and stability [18], [19], operation under short-circuit and avalanche conditions [20] and power cycling capability [21], [22]. Countermeasures to some of these problems can already be found in the literature and involve improved semiconductor manufacturing processes and advanced packaging technologies.
A final possibility considered here to improve PV yield is the use of multilevel inverters. The associated advantages are similar to that of WBG devices, with exception to the possibility of simplifying the inverter system. In reality, the implementation of multilevel inverters often comes with the price of increasing semiconductor switch part count and complexity.
Having the chart of Figure 1.3 in mind, this Ph.D. project aims to redesign, based on WBG semiconductors, the standard DC-AC converter architecture in PV applications. With the purpose of suppressing the DC-link capacitors and simplifying the inverter topology, the single-stage three-phase Current Source Inverter (CSI) is proposed. Thanks to their compatible voltage rating levels together with exceptional on-state characteristics, Silicon Carbide (SiC) devices are employed in this work. Thereby, implementing the CSI topology with SiC devices brings the possibility to increase the grid voltage level, suppress the DC-link capacitors, reduce the switches part count and simplify the power stage topology. Each of these advantages and potentials limitations will be discussed in detail in the following chapters.

READ  Spawning and non-breeding activity of adult giant bullfrogs

Review of Three-phase Photovoltaic Inverter Topologies

PV inverters are normally voltage source-based converters [1], [2] and can be classified with regard to their number of power processing stages. Thereby, two main categories arise, namely single-stage and multistage inverters [2]. In Figure 2.1 (a) and (b) are depicted, respectively, the block diagram of a three-phase single- and two-stage inverter topologies.
Figure 2. 1 – Three-phase inverter block diagram for a) single-stage and b) two-stage architectures. c) Single-stage three-phase NPC inverter. Two-stage three-phase architectures: d) boost + VSI, e) symmetric boost + 5-level VSI, f) symmetric boost + 4-level VSI and g) symmetric boost + 5-level TCC VSI.

Table of contents :

Chapter 1
Introduction
1.1 Photovoltaic Scenario Overview
1.2 Power Electronics for Photovoltaics
1.3 Research Motivations and Objectives
1.4 Dissertation Outline
1.5 References
Chapter 2
State of the Art Review
2.1 Review of Three-phase Photovoltaic Inverter Topologies
2.2 SiC vs. Si – Potentials of Wide-bandgap Devices for Performance Improvement in Energy Conversion Systems
2.2.1 Brief Overview of WBG Materials and Properties
2.2.2 SiC-based Semiconductor Devices
2.2.3 SiC Devices for Inverter Performance Improvement
2.3 Three-phase Current Source Inverter for Photovoltaic Applications: a Single-stage Topology as an Alternative Solution to Voltage Source-based Converters
2.4 Research Challenges and Opportunities
2.5 References
Chapter 3
Three-phase Current Source Inverter Topology
3.1 Introduction
3.2 Principle of Operation
3.2.1 Space Vector Transformation
3.2.2 CSI Operation in Combination with a PV String – Steady-state Analysis
3.2.3 Analysis of the Switching Cell Disposition
3.2.4 Current and Voltage Stresses of Power Semiconductors
3.2.5 Typical waveforms
3.3 Semiconductor Conduction Losses
3.3.1 Static modeling of semiconductor devices
3.3.2 Conduction losses estimation
3.3.3 Comparison of different CSI switch configurations in terms of conduction losses
3.4 Switching Event in the CSI Topology
3.4.1 Simulation-based comparison of switching event in the CSI and VSI topologies
3.5 Space Vector Modulation Strategies
3.5.1 Methodology for Current Spectrum Determination
3.5.2 Calculation Results
3.6 Summary and Conclusion
3.7 References
Chapter 4
Design and Characterization of a Novel 1.7kV Full-SiC Power Module
4.1 Introduction
4.2 Power Integration for Current Source-based Topologies
4.2.1 Module Substrate Design
4.2.2 Electrical Performance Evaluation
4.2.3 Qualitative Modeling of a Coupled Microstrip – a Generic Case
4.3 Industrial Realization of the Full-SiC CSI Module
4.3.1 Material Selection
4.3.2 Definition of Module Voltage and Current Ratings
4.3.3 Analysis of the Packaging Thermal Resistance Sharing
4.3.4 Module Design
4.4 Implementation of the Double-Pulse Test Characterization Method for SiC Devices Benchmarking
4.4.1 Switching Losses Measurement Methods
4.4.2 Double-Pulse Test Method
4.4.3 Current Measurement Techniques
4.4.4 Selection of SiC Devices
4.5 Dynamic Characterization of the Developed CSI Module
4.5.1 Switching Speed Characterization
4.5.2 Switching Losses Characterization
4.5.3 Influence of Probes’ Relative Delay on Switching Losses
4.5.5 Decoupling Capacitor Current Measurement and Sizing
4.6 Summary and Conclusion
4.7 References
Chapter 5
Design and Experimental Evaluation of the 60kW CSI Prototype 
5.1 Introduction
5.2 Sizing of Passive Components
5.2.1 DC Inductor
5.2.2 DC Capacitor
5.2.3 AC Filter
5.3 Thermal Management
5.3.1 Heat Sink Sizing
5.3.2 SiC Module Thermal Runaway
5.3.3 Experimental Evaluation of the Module Thermal Behavior
5.4 Efficiency Estimation
5.4.1 CSI Operation on a Three-phase 800V Grid
5.5 Comparison with a Standard Solution: DC-DC Boost Converter in combination with a Three-phase Two-level Voltage Source Inverter
5.6 Hardware Description
5.6.1 Protection Circuit
5.6.2 Gate Driver
5.6.3 Busbar
5.6.4 Setup for Direct Measurement of Junction Temperatures
5.7 CSI Characterization
5.7.1 Heat Sink Characterization for Calorimetric Semiconductor Losses Determination
5.7.2 Waveforms at Nominal Switching Operation
5.7.3 Influence of Switching Frequency on Semiconductor Losses
5.7.4 Efficiency Measurements
5.8 Summary and Conclusion
5.9 References
Chapter 6
Conclusions and Future Work
6.1 Power Integration for CSI Applications
6.2 Dynamic Characterization of 1.7kV SiC Devices
6.3 Inverter Level
6.4 References

GET THE COMPLETE PROJECT

Related Posts