Probabilistic Analysis of Power Network Susceptibility to GICs

TL;DR

This paper introduces a probabilistic method to analyze power network susceptibility to geomagnetically induced currents (GICs), enabling empirical risk assessment without detailed geophysical or network data.

Contribution

A novel probabilistic engineering approach that estimates GIC impact using measured geomagnetic data, bypassing complex geophysical modeling and network topology requirements.

Findings

Empirical ensembles effectively assess GIC exposure during geomagnetic storms.

Nodes in the TVA network are ranked by susceptibility to GICs.

Calibrated models align with existing extreme value analyses.

Abstract

As reliance on power networks has increased over the last century, the risk of damage from geomagnetically induced currents (GICs) has become a concern to utilities. The current state of the art in GIC modelling requires significant geophysical modelling and a theoretically derived network response, but has limited empirical validation. In this work, we introduce a probabilistic engineering step between the measured geomagnetic field and GICs, without needing data about the power system topology or the ground conductivity profiles. The resulting empirical ensembles are used to analyse the TVA network (south-eastern USA) in terms of peak and cumulative exposure to 5 moderate to intense geomagnetic storms. Multiple nodes are ranked according to susceptibility and the measured response of the total TVA network is further calibrated to existing extreme value models. The probabilistic engineering step presented can complement present approaches, being particularly useful for risk assessment of existing transformers and power systems.