Session: 01-04-02 Design & Analysis-II

Paper Number: 126505

126505 - Design and Optimisation of Parallel Hybrid Converter Connected With Lcl Filter for Offshore Wind System 

The growth of industrial high-power electronics is expected to accelerate significantly, driven by the development of renewable energy sources such as wind and solar power. A common converter topology for multi-megawatt DC-AC applications, including grid-interfacing converters for renewable energy, uninterruptible power supplies (UPS), and battery storage, is the two-level converter with an LCL filter output, typically operated at a fixed switching frequency. Currently, Silicon IGBTs are the most widely used semiconductor devices for medium to high-power conversion. However, their switching frequency in high-power applications is typically limited to 2-3 kHz due to relatively high switching losses.

Wide-bandgap power semiconductors like Silicon Carbide (SiC) and Gallium Nitride (GaN) offer several advantages, including higher switching efficiency, increased bandwidth, and a smaller form factor when compared to traditional Silicon power semiconductors. This provides greater flexibility in dynamic control systems and allows for the design of more efficient power converters, resulting in improved system robustness and power quality. SiC MOSFET converters have demonstrated significant performance improvements over previous Si IGBT-based converter topologies. Nevertheless, the high cost of wide-bandgap semiconductors and the limited availability of high-current modules remain barriers to their widespread adoption in high-power applications. SiC MOSFETs, for instance, are typically 2-4 times more expensive than Si MOSFETs and 4-6 times more expensive than Si IGBTs, according to the Digikey official website. In commercial applications, wide-bandgap semiconductors are only available with current ratings up to 120A, which restricts their direct replacement of Si IGBTs in high-power applications.

To harness the advantages of wide-bandgap devices while addressing the cost issue, a hybrid approach combining wide-bandgap converters with Si converters is a viable alternative. This hybrid topology leverages the strengths of both converters and balances the high initial cost of wide-bandgap devices with long-term savings in switching costs.

The role of filters between the converter and the grid network is also critical. These filters are essential for mitigating current ripples and achieving the desired output current with limited Total Harmonic Distortion (THD) of up to 5%. Various types of filters, such as L-filters, CL-filters, and LCL-filters, are commonly used to reduce harmonic distortion in connection topologies. While L and CL filters are popular due to their simplicity and fewer components, they exhibit slower dynamic responses in high-power applications, and L-inductors may experience high voltage drops in the output system.

To improve the output performance, the LCL filter is a common choice to replace traditional L-filters. It improves the output current from a voltage source converter and reduces switching harmonics. However, designing an LCL filter is more complex due to the increased number of control variables, making research into LCL filter connections and comprehensive analyses of this complex topology essential.

This paper focuses on designing a Parallel Hybrid Converter (PHC) topology with an LCL filter. It discusses the influence of different impedance and capacitors of the PHC's output filter on the switching frequency of IGBT and MOSFET converters and the THD of the output grid current. The paper also introduces and utilizes the Finite-Set Model Predictive Controller (FS-MPC) for control. Additionally, it analyses the static and dynamic performance, including system faults and harmonic injection, based on real-time FPGA circuitry for a 90kVA converter to validate simulation results and practical system performance. The study of PHC enables the efficient conversion of variable offshore wind and tidal energy into the onshore grid while maintaining optimal offshore grid synchronization, grid code compliance and power quality.

Presenting Author: Ning Li The University of Edinburgh

Presenting Author Biography: Ning Li (Student Member, IEEE) received the B.Sc. in electrical engineering and automation in 2018 from the South China University of Technology, Guangdong, China, and the M.Sc. degree in sustainable energy systems in 2019 from the University of Edinburgh, Scotland, U.K. where he is currently working toward the Ph.D. degree in power electronics within the Institute for Energy Systems, School of Engineering. His current research interests include model predictive, learning-based control strategies and optimization of power electronics converters.

Authors:

Ning Li The University of Edinburgh
Stephen Finney The University of Edinburgh
Paul Judge The University of Edinburgh
Yixiao Zhang Nanyang Technological University
Eddie Yin Kwee Ng Nanyang Technological University