Proactive Islanding of the Power Grid to Mitigate High-Impact Low-Frequency Events

TL;DR

This paper introduces a proactive islanding methodology for power grids that enhances resilience against high-impact, low-frequency events by partitioning the network into self-sustaining islands using spectral clustering and power flow validation.

Contribution

It presents a novel spectral clustering-based approach for proactive grid islanding, validated with realistic PJM network models to improve system resilience during extreme events.

Findings

Islanding reduces cascade propagation during outages

Power flow analysis confirms island viability

Seasonal variations affect island configurations

Abstract

This paper proposes a methodology for enhancing power systems resiliency by proactively splitting an interconnected grid into small self-sustaining islands in preparation for extreme events. The idea is to posture the system so that cascading outages can be bound within affected areas, preventing the propagation of disturbances to the rest of the system. This mitigation strategy will prove especially useful when advance notification of a threat is available but its nature not well understood. In our method, islands are determined using a constrained hierarchical spectral clustering technique. We further check the viability of the resultant islands using steady-state AC power flow. The performance of the approach is illustrated using a detailed PSS/E model of the heavily meshed transmission network operated by PJM Interconnection in the eastern USA. Representative cases from different seasons show that variations in power flow patterns influence island configuration.