Implementation of aggregated Distribution System Models and application for design of efficient topology and control

Abstract

This thesis aims to utilise Open Distribution System Model (ODSM) based on real-world grid data. For these purposes, model had been established from results of master project, that had been aimed to collect representative data for further model creation. The model was assembled from various independent grid models data and recreated as Open Distribution System Model. The intention of the thesis is to build a model in open-source format where real grids topologies, loads and generation data were combined and efficiently anonymised. Results of the master projects provided representative medium voltage (MV) grids from Distribution System Operator (DSO) region of operation and helped to established highly detailed distribution system with fully functional real-life topological and electrical behaviour. The ODSM was developed to align with current and future requirements for resilient energy system operation. For efficiency and fast computation of various grid analysis, the original detailed Open Distribution System Model data was aggregated and downsized for use in educational and research settings. Therefore, the aggregation algorithm allows to change scale of aggregation which results in different final model sizes. For improved aggregation, the model was built and validated with actual DSO data and full-size Open Distribution System Model against aggregated. To this end, an aggregation algorithm was utilized, and a comparison between aggregated and reference data was conducted. To evaluate the aggregation, an additional crosscheck between each main electrical characteristics and topology was performed. The result of this process is detailed Open Distribution System Model and aggregated model with limited number of nodes and lines, which presents actual parts of the existing distribution network. The model with a modified topology of primary substations in the medium-voltage (MV) distribution grid and a manually constructed high-voltage (HV) distribution grid with interconnection to a fraction of manually constructed transmission system (EHV) represented as feeding node with extra-long transmission lines loop connected through distribution system. This model configuration allows future expansion to support more complex system calculation, optimisation and analysis, including the interconnection of additional system operators and the study of system states under various operational and maintenance scenarios. Although the present version is limited in scope, the model is architected for expansion. As it matures, it is expected to enable more complex calculations spanning probabilistic and scenario-based analyses, multi-scale co-simulation and digital-twin integration covering hybrid load forecasting, decentralised generation modelling, hosting-capacity assessment, optimisation strategies, system flexibility, energy-supply optimisation, and advanced stability, transient, and fault studies, alongside time-of-use and load-profile derivation.

Date of Award	2025
Original language	English
Supervisor	Andreas Abart (Supervisor)