LOW-COST AND HIGH-RELIABILITY, SINGLE-STAGE MICROINVERTER FOR RESIDENTIAL SOLAR APPLICATIONS

Abstract

Recent attention has focused on reduced cost, improved reliability, and enhanced functionality of power electronics in residential solar applications, specifically in the microinverter architecture. In particular, the Department of Energy SunShot Initiative 2030 goals target a Levelized Cost of Electricity (LCOE) of less than 5 ¢/kWh. Traditional microinverter costs represent nearly 22 % of fixed system-level costs and approximately 45 % of variable operation and maintenance costs related to reliability, inhibiting achievable LCOE. This dissertation proposes a next generation microinverter as a key enabler for advancements required to realize the 2030 goals, from cost and reliability perspectives. The proposed microinverter circuit utilizes a single-stage topology based on the Dual Active Bridge (DAB) circuit, leveraging a combination of wide-bandgap Gallium Nitride (GaN) devices and Si devices at high switching frequency. Additionally, the proposed converter realizes low-frequency energy storage through an active power decoupling (APD) circuit, which enables the use of high-reliability film capacitors in contrast to the traditional use of lower-reliability electrolytic capacitors. Aside from topological advantages such as inherent galvanic isolation and low component count, design decisions and component selections are supported by multi-objective optimization, pushing the boundaries in the converter design. The proposed microinverter circuit is analyzed in steady-state operation, where an improved analytical modelling technique is required to properly predict converter performance, and enable control-level optimization. Main-circuit parametric design optimization is performed to select transformer turns ratio and leakage inductances which minimize the converter’s efficiency drop due to conduction loss, while enabling near-uniform zero-voltage-switching transitions minimizing switching-related losses. The analysis is extended to multi-objective analysis targeting minimization of cost, area, and efficiency drop, informing design tradeoffs and component selection both at the topology and device levels. To improve the performance of the main-circuit transformer, and reduce the cost with a planar PCB-based design, a novel integrated-leakage transformer is developed which adopts performance advantages over similar state-of-the-art designs. Finally, holistic multi-objective optimization procedure is developed for the APD based on cost, efficiency, and power density, to enable holistic component selection. To validate the design and associated analyses, a proof-of-concept for the main-circuit and APD are designed and tested.