Comparative Life Cycle Cost Analysis of Si and SiC PV Converter Systems Based on Advanced $\eta$- $\rho$-$\sigma$ Multiobjective Optimization Techniques

Abstract

This paper presents a novel virtual prototyping routine for power electronic converter systems. The approach facilitates a comprehensive and systematic benchmarking of different converter concepts based on a multiobjective optimization regarding the efficiency, power density, and costs. The underlying modeling framework is based on detailed and experimentally verified models. In particular, novel cost data as well as unpublished switching loss and core loss measurements are incorporated. The proposed virtual prototyping routine is employed to carry out a comparative study of the potential of Si and SiC semiconductors in a 10-kW residential three-phase photovoltaic inverter application. For this purpose, a state-of-the-art hard-switched three-level Si insulated-gate bipolar transistor (IGBT) system is compared to a hard-switched and to a soft-switched two-level SiC MOSFET system. The candidate systems for each concept are selected among the $\boldsymbol {\eta}$ - $\boldsymbol {\rho}$ - $\boldsymbol {\sigma}$ Pareto-optimized designs based on the life cycle costs. The hard-switched two-level SiC candidate system is found to be the most attractive solution featuring the lowest life cycle costs. When compared to the Si-based candidate system, not only a better power density and efficiency result. At the same time, besides the lower life cycle costs ( $-$ 22%), lower component costs ( $-$ 5%) can also be attained. The attractiveness of the found SiC solution is underlined by the simple control and the lowest component count among all concepts.