Assessment of Graph-Based Clustering Algorithms for Alternative Bidding Zone Configurations

This paper focuses on model-based algorithms to be applied for identifying alternative configurations in the context of a bidding zone review process for the European Day-Ahead Electricity Market operation that adopts a zonal model to manage the transmission grid congestions. In particular, the paper presents evolved graph-based clustering algorithms designed to define bidding zones formed only by interconnected nodes. Then, a method that evaluates the performance of the resulting alternative bidding zone configurations by calculating specific performance indicators is proposed. The clustering algorithms proposed are compared with other algorithms previously developed by the authors in their application to a large-scale model and extended data set related to the Italian Transmission Network, and relevant results are presented.