Identification of typical urban grid structures across voltage levels by clustering distribution grid big data

摘要

With the background of changing scenarios for power grids in terms of new loads and distributed generation, grid planning is repeatedly confronted with the task of selecting representative grids, planning them and deriving recommendations for measures from them and applying the findings on the entire distribution grid. Therefore, the aim of this contribution is to identify representative urban distribution grids from six distribution system operators by applying a clustering procedure for more than 7,000 low-voltage and more than 140 medium-voltage grids. In addition, a cross-voltage level analysis is carried out and validated with market data and geodata, as well as evaluated in detail for two urban grid areas and visualised in a geoinformation system. It is proven that the grid areas are basically very similar in terms of building structure and grid structure.