A framework for testing stealthy attacks in energy grids

Abstract

The progressive integration of traditional Information and Communication Technologies (ICT) hardware and software into the supervisory control of modern Power Grids (PG) has given birth to a unique technological ecosystem. Modern ICT handles a wide variety of advantageous services in PG, but in turn exposes PG to significant cyber threats. To ensure security, PG use various anomaly detection modules to detect the malicious effects of cyber attacks. In many reported cases the newly appeared targeted cyber-physical attacks can remain stealthy even in presence of anomaly detection systems. In this paper we present a framework for elaborating stealthy attacks against the critical infrastructure of power grids. Using the proposed framework, experts can verify the effectiveness of the applied anomaly detection systems (ADS) either in real or simulated environments. The novelty of the technique relies in the fact that the developed “smart” power grid cyber attack (SPGCA) first reveals the devices which can be compromised causing only a limited effect observed by ADS and PG operators. Compromising low impact devices first conducts the PG to a more sensitive and near unstable state, which leads to high damages when the attacker at last compromises high impact devices, e.g. breaking high demand power lines to cause blackout. The presented technique should be used to strengthen the deployment of ADS and to define various security zones to defend PG against such intelligent cyber attacks. Experimental results based on the IEEE 14-bus electricity grid model demonstrate the effectiveness of the framework.