Topological analyses are crucial when assessing network robustness or generating synthetic power grids. We analyze the topology of fifteen European transmission networks by using complex-network metrics. The study includes two voltage levels: 400 kV and 200 kV. We study these levels both independently and as a single combined grid. Degree distribution, characteristic path length, network diameter, betweenness centrality and global clustering coefficient are explored in order to understand network topology and to explain observed differences among countries. We analyze empirically whether those metrics scale or not with network size. Our main conclusions are the following. The number of lines scales linearly with the number of substations in power networks. Characteristic path length and network diameter grow logarithmically with respect to the number of nodes. Finally, mean value and maximum value of betweenness centrality follow a power-law with respect to number of nodes. The clustering coefficient does not scale with network size and varies widely among countries. In addition, we compare global clustering coefficient and characteristic path length with their typical values in random networks. Whilst global clustering coefficient is larger in power networks (compared to random networks), characteristic path length is similar in both types of graphs. Consequently, the small-world index is higher than one for most power networks. However, there are some exceptions when considering the 400 kV and 220 kV layers as independent networks. Our conclusions improve the current understanding of power network topology, which is essential for generating synthetic power grids and in the assessment of network robustness.
Registration date: 2018-02-26