Phase selection algorithms to minimize cost and imbalance in U.S. synthetic distribution systems

Abstract

The increasing penetration of distributed energy resources (DERs) is necessitating development of new algorithms and tools for distribution system operation and planning; however, there is a lack of publicly available electrical distribution test systems at the large scale-multi-feeder, multi-substation- which are required for realistically evaluating the performance and scalability of these new developments. This paper presents a Reference Network Model (RNM) aimed to design large-scale, U.S.-style, synthetic distribution systems. Special emphasis is placed on two algorithms that allow multiphase feeder design: 1) a method to select the most suitable number of phases for each section considering the connected customers, and 2) a method to assign phases to the users to provide near-balanced phasing while maintaining realistic levels of imbalance. The performance of the developed algorithms is verified by comparing the obtained system designs with the original IEEE 8,500-node test feeder and the Electric Power Research Institute (EPRI) J1 feeder.

The increasing penetration of distributed energy resources (DERs) is necessitating development of new algorithms and tools for distribution system operation and planning; however, there is a lack of publicly available electrical distribution test systems at the large scale-multi-feeder, multi-substation- which are required for realistically evaluating the performance and scalability of these new developments. This paper presents a Reference Network Model (RNM) aimed to design large-scale, U.S.-style, synthetic distribution systems. Special emphasis is placed on two algorithms that allow multiphase feeder design: 1) a method to select the most suitable number of phases for each section considering the connected customers, and 2) a method to assign phases to the users to provide near-balanced phasing while maintaining realistic levels of imbalance. The performance of the developed algorithms is verified by comparing the obtained system designs with the original IEEE 8,500-node test feeder and the Electric Power Research Institute (EPRI) J1 feeder.