Ramp-Induced Data Attacks on Look-Ahead Dispatch in Real-Time Power Markets

Abstract

This paper presents a new class of false data injection attacks on state estimation, which may lead to financial arbitrage in real-time power markets with an emerging look-ahead dispatch model. In comparison with prior work of cyber attack on static dispatch where no inter-temporal ramping constraint is considered, we propose a novel attack strategy with which the attacker can manipulate, in look-ahead dispatch, the limits of ramp constraints of generators. It is demonstrated that the proposed attack may lead to financial profits via malicious capacity withholding of selected generators, while being undetected by the existing bad data detection algorithm embedded in the state estimator. The feasibility of such cyber attacks and their economic impact on real-time electricity market operations are illustrated in the IEEE 14-bus system.