Analysis of major failures in Europe's power grid

TL;DR

This study analyzes seven years of Europe's power grid failure data, fitting probability distributions to understand the system's failure dynamics, revealing that while power law models are somewhat supported, other factors likely influence failures.

Contribution

It provides an empirical analysis of power grid failures over seven years and evaluates the suitability of power law models for the failure distribution.

Findings

Power law models have moderate support in the data.

Other causal factors likely influence power grid failures.

The tail of the distribution contains limited data points.

Abstract

Power grids are prone to failure. Time series of reliability measures such as total power loss or energy not supplied can give significant account of the underlying dynamical behavior of these systems, specially when the resulting probability distributions present remarkable features such as an algebraic tail, for example. In this paper, seven years (from 2002 to 2008) of Europe's transport of electricity network failure events have been analyzed and the best fit for this empirical data probability distribution is presented. With the actual span of available data and although there exists a moderate support for the power law model, the relatively small amount of events contained in the function's tail suggests that other causal factors might be significantly ruling the system's dynamics.