Visible to the public Load Redistribution Attacks in Multi-Terminal DC Grids

TitleLoad Redistribution Attacks in Multi-Terminal DC Grids
Publication TypeConference Paper
Year of Publication2022
AuthorsZhang, Zhi Jin, Bloch, Matthieu, Saeedifard, Maryam
Conference Name2022 IEEE Energy Conversion Congress and Exposition (ECCE)
Date Publishedoct
Keywordscommand injection attacks, composability, cyber-attack, cyber-physical system, cyber-security, cyberattack, False Data Injection, Lead, Metrics, multi-terminal dc grid, Optimization, power grids, pubcrawl, reliability, resilience, Resiliency, Systems operation, Voltage control, voltage-sourced converter, vulnerability analysis
AbstractThe modernization of legacy power grids relies on the prevalence of information technology (IT). While the benefits are multi-fold and include increased reliability, more accurate monitoring, etc., the reliance on IT increases the attack surface of power grids by making them vulnerable to cyber-attacks. One of the modernization paths is the emergence of multi-terminal dc systems that offer numerous advantages over traditional ac systems. Therefore, cyber-security issues surrounding dc networks need to be investigated. Contributing to this effort, a class of false data injection attacks, called load redistribution (LR) attacks, that targets dc grids is proposed. These attacks aim to compromise the system load data and lead the system operator to dispatch incorrect power flow commands that lead to adverse consequences. Although similar attacks have been recently studied for ac systems, their feasibility in the converter-based dc grids has yet to be demonstrated. Such an attack assessment is necessary because the dc grids have a much smaller control timescale and are more dependent on IT than their traditional ac counterparts. Hence, this work formulates and evaluates dc grid LR attacks by incorporating voltage-sourced converter (VSC) control strategies that appropriately delineate dc system operations. The proposed attack strategy is solved with Gurobi, and the results show that both control and system conditions can affect the success of an LR attack.
NotesISSN: 2329-3748
DOI10.1109/ECCE50734.2022.9948106
Citation Keyzhang_load_2022