A Framework for Generating Synthetic Distribution Feeders using OpenStreetMap

TL;DR

This paper presents a framework that generates realistic synthetic distribution feeders mapped to actual geographic locations using OpenStreetMap data, aiding power systems research with extensive, geographically accurate test cases.

Contribution

The authors introduce a novel framework that creates synthetic distribution feeders with minimal data, integrating real geographic and population data for widespread US application.

Findings

Generated thousands of realistic distribution feeders

Able to produce feeders for entire ZIP codes

Outputs compatible with OpenDSS for simulation

Abstract

This work proposes a framework to generate synthetic distribution feeders mapped to real geo-spatial topologies using available OpenStreetMap data. The synthetic power networks can facilitate power systems research and development by providing thousands of realistic use cases. The location of substations is taken from recent efforts to develop synthetic transmission test cases, with underlying real and reactive power in the distribution network assigned using population information gathered from United States 2010 Census block data. The methods illustrate how to create individual synthetic distribution feeders, and groups of feeders across entire ZIP Code, with minimal input data for any location in the United States. The framework also has the capability to output data in \OpenDSS format to allow further simulation and analysis.