New Frontiers for Blackout Prevention: A Study of the Vulnerability Frontier and its Use for Contingency Analysis and Reliability Assessment
Abstract
From infrastructure to healthcare to national security, global dependence on bulk power systems has never been more critical-placing reliability at a premium. Disruptions to power are consequential, and data on the frequency and size of blackouts over the past several decades is troubling. The existence of a power law in the frequency distribution of blackout size suggests the trends are no anomaly and more should be done in preparation for impending severe outages. The vulnerability frontier, defined as the set of points relating the maximum power disruption as a function of the number of lines removed from service, offers a unique screening approach to help examine certain worst-case events. To quantify grid reliability, associated scalar metrics influenced both by network topology and the pattern of power injections over the network are also framed. Existing contingency analysis practices typically rely on user-specified lists due to computational barriers faced when considering anything greater than N-1 or N-2 events. Thus, an opportunity is present for the frontier to better inform these lists with insightful selections of contingencies to assess.
The vulnerability frontier is studied on a 7977-bus synthetic grid model of the Midwest transmission network. A series of test scenarios are posed to probe the response of the frontier under seasonal and daily load profiles, transmission and generation outages, as well as in instances with increased renewable generation. These observational case studies show the effectiveness of the frontier in capturing grid weaknesses and lend a basis for further study. Modifications to the frontier formulation are provided in an effort to highlight cascading outage events in particular. Results are validated following a comparison to a modified version of the well-known Oak Ridge-PSERC-Alaska (OPA) blackout model.
Permanent Link
http://digital.library.wisc.edu/1793/79126Type
Thesis
Part of
Licensed under: