Decreased resilience in power grids under dynamically induced vulnerabilities

TL;DR

This paper presents a new methodology inspired by percolation theory to assess power grid resilience, revealing that targeting dynamical vulnerabilities significantly increases blackout risk, especially in homogeneously powered networks.

Contribution

It introduces a combined structural and dynamical vulnerability measure for power grids and applies it to real and synthetic networks to analyze resilience.

Findings

Targeted attacks on dynamical vulnerabilities increase blackout likelihood.

Homogeneous power distribution makes nodes more susceptible to failures.

Power grid resilience decreases under dynamical vulnerability-focused attacks.

Abstract

In this paper, a methodology inspired on bond and site percolation methods is applied to the estimation of the resilience against failures in power grids. Our approach includes vulnerability measures with both dynamical and structural foundations as an attempt to find more insights about the relationships between topology and dynamics in the second-order Kuramoto model on complex networks. As test cases for numerical simulations, we use the real-world topology of the Colombian power transmission system, as well as randomly generated networks with spatial embedding. It is observed that, by focusing the attacks on those dynamical vulnerabilities, the power grid becomes, in general, more prone to reach a state of total blackout, which in the case of node removal procedures it is conditioned by the homogeneity of power distribution in the network.