Node-­Failure and Islanding in National Grid Scale Electricity Distribution Networks

Ken A. Hawick


Islanding, network fragmentation, electrical power distribution, complex networks, risk analysis


Electrical power distribution networks are becoming more complex with increased variability in the type and properties of power station generators. National grid scale distribution networks are governed by many physical and geographic constraints and it is difficult to identify points of vulnerability. We employ some static graph analysis metrics to study how an approximation of the UK National Grid islands first into regions then small clusters of generators and substations. We use a progressive node culling procedure based upon the between-ness centrality metric to investigate numerically the properties of the overall network and individual islands as nodes and hence distribution lines fail. We find that the UK Grid system breaks first into regional islands, but is remarkably stable against node failures, breaking further into islands that are potentially viable at some service provision level. The between-ness metric identifies the most critical nodes in the network.

Important Links:

Go Back