Co-optimizing High and Low Voltage Systems: Bi-Level vs. Single-Level Approach

Abstract

This paper presents a bi-level optimization framework applied to optimize system performance with (i) increasing presence of distributed energy resources (DER) at the low-voltage level, and (ii) variable wind power generation at the high-voltage level. The paper investigates various system configurations with increasing presence of microgrids, with active devices. System simulations quantify system performance in terms of cost, first using the traditional single-level optimization framework, and second using the proposed bi-level framework. Comparisons between the system with traditional, passive distribution systems and with microgrids are also presented, with results again quantified via the interconnected system operating costs. Results show that at low levels of DER and microgrid penetration, traditional (single-level) system optimization algorithms perform adequately as compared to the proposed bi-level optimization framework. However, as DER and microgrid penetration increase, the traditional single-level framework does not accurately capture the full system benefits of distributed technologies. The results demonstrate that new optimization algorithms, such as the proposed bi-level framework, will be required if the benefits of DER are to be accurately quantified in the evolving power system.