Search for a command to run...
Open Distribution System Model (ODSM) enables reproducible studies of distribution planning and operation, but utility-grade datasets are rarely shareable due to confidentiality, heterogeneity across grid projects, and excessive model details. This paper documents the creation workflow of an anonymized ODSM derived from a real distribution system. First, statistical feature sets are extracted from a large pool of medium-voltage (MV) and low-voltage (LV) grid models. Second, representative MV and LV subsets are selected using a k-d tree-guided nearest-neighbor search with an RMS-based objective and a stopping criterion. Third, the selected projects are programmatically assembled in DIgSILENT PowerFactory via Python, harmonizing libraries, naming, and georeferencing while preserving topology and steady-state behavior. Finally, a multi-stage aggregation algorithm reduces node and line counts for fast studies while maintaining electrical characteristics. The resulting full-size and aggregated ODSM variants are suited for education and research and provide a consistent basis for further chapters on advanced studies and applications.