Significance of Grid Level Scope and Reduction Methods for Effective Congestion Management Modeling

This paper investigates the influence of grid reduction methods on congestion management modeling in Germany. Using PyPSA-Eur, 15 grid configurations were created across three scenario levels: (i) inclusion or exclusion of the highvoltage (HV) grid, (ii) aggregation of HV into the extra-high voltage (EHV) level, and (iii) application of spatial grid clustering algorithms. All scenarios were calibrated for 2019 to allow validation against historical congestion management data. The evaluation combines topological indicators with redispatch volumes and their regional distribution. Results show that a fullresolution HV grid without reduction leads to unrealistic redispatch requirements. In contrast, both HV reduction and spatial clustering applied to the HV grid produce more plausible outcomes and more accurately reflect the spatial distribution of congestion management, underscoring the importance of a highresolution grid level. Since all spatial clustering algorithms consider connectivity and apply consistent clustering features, differences between them remain minor. Despite the improvements achieved by combining grid level scope and reduction methods, all scenarios still overestimate redispatch compared to historical data, with the closest match exceeding by a factor of 2.26. These results underscore the necessity of grid reduction for realistic congestion management modeling and highlight the inherent limitations of grid models