Load dependence of power outage statistics

Biswas, S. and Goehring, L. ORCID: 0000-0002-3858-7295, 2019. Load dependence of power outage statistics. EPL (Europhysics Letters), 126 (4): 44002. ISSN 1286-4854

[img]
Preview
Text
14060_Goehring.pdf - Post-print

Download (3MB) | Preview

Abstract

Dynamics of power outages remain an unpredictable hazard in spite of expensive consequences. While the operations of the components of power grids are well understood, the emergent complexity due to their interconnections give rise to intermittent outages, and power-law statistics. Here we demonstrate that there are additional patterns in the outage size distributions that indicate the proximity of a grid to a catastrophic failure point. Specifically, analysis of the data for the U.S. between 2002-2017 shows a significant anti-correlation between the exponent value of the power-law outage size distribution and the load carried by the grid. The observation is surprisingly similar to dependencies noted for failure dynamics in other multi-component complex systems such as sheared granulates and earthquake statistics, albeit under much different physical conditions. This inspires a generic threshold-activated model, simulated in realistic network topologies, which could successfully reproduce the exponent variation in similar range. Given sufficient data, the methods proposed here can be used to indicate proximity to failure points and forecast probabilities of major blackouts with a non-intrusive measurement of intermittent grid outages.

Item Type: Journal article
Publication Title: EPL (Europhysics Letters)
Creators: Biswas, S. and Goehring, L.
Publisher: EDP Sciences
Date: 26 June 2019
Volume: 126
Number: 4
ISSN: 1286-4854
Identifiers:
NumberType
10.1209/0295-5075/126/44002DOI
Divisions: Schools > School of Science and Technology
Depositing User: Jonathan Gallacher
Date Added: 13 Jun 2019 08:34
Last Modified: 04 Jul 2019 14:20
URI: http://irep.ntu.ac.uk/id/eprint/36749

Actions (login required)

Edit View Edit View

Views

Views per month over past year

Downloads

Downloads per month over past year