Assessing and managing the information and communication risk of power generation

We describe a model-based assessment of information and communication technology (ICT) risk that produces statistical samples by simulating the attacks of intelligent agents. To support this assessment, we have developed an integrated set of tools, the Haruspex suite. Some of its tools build the models of the target system and those of the agents that other tools apply to simulate the agent attacks. Further tools analyze the output of the simulation. After outlining the proposed approach and the suite, we describe the assessments of two industrial control systems that supervise, respectively, a thermoelectric generation plan and a hydroelectric one. To simplify the presentation of the output of these assessments, we introduce the security stress, a synthetic measure of how a system resists to attacks.