Nikolaus Wirtz

Modern critical infrastructures (e.g. Critical Energy Infrastructures) are increasingly evolving into complex and distributed networks of Cyber-Physical Systems. Although the cyber systems provide great flexibility in the operation of critical infrastructure, it also introduces additional security threats that need to be properly addressed during the design and development phase. In this landscape, resilience and robustness by design are becoming fundamental requirements. In order to achieve that, new approaches and technological solutions have to be developed that guarantee i) the fast incident/attack detection; and ii) the adoption of proper mitigation strategies that ensure the continuity of service from the infrastructure. The “Double Virtualization” emerged recently as a potential strategy/approach to ensure the robust and resilient design and management of critical energy infrastructures based on Cyber-Physical Systems. The presented approach exploits the separation of the virtual capabilities/functionalities of a device from the physical system and/or platform used to run/execute them while allowing to dynamically (re-) configure the system in the presence of predicted and unpredicted incidents/accidents. Internet-based technologies are used for developing and deploying the envisioned approach.

This paper introduces a flexible framework to analyze cascading effects in the interdependent power and information and communications technology (ICT) networks that that comprise a power system. This framework supports integration of interdependencies between the power grid and various ICT networks, but also of domain-specific intra-dependencies of these different subsystems. The framework is applied to model a simple example system, where three failure scenarios are defined and simulated to showcase the applicability of the framework for the investigation of cascading effects.

With the increasing interdependence of critical infrastructures, the probability of a specific infrastructure to experience a complex cyber-physical attack is increasing. Thus it is important to analyze the risk of an attack and the dynamics of its propagation in order to design and deploy appropriate countermeasures. The attack trees, commonly adopted to this aim, have inherent shortcomings in representing interdependent, concurrent and sequential attacks. To overcome this, the work presented here proposes a stochastic methodology using Petri Nets and Continuous Time Markov Chain (CTMC) to analyze the attacks, considering the individual attack occurrence probabilities and their stochastic propagation times. A procedure to convert a basic attack tree into an equivalent CTMC is presented. The proposed method is applied in a case study to calculate the different attack propagation characteristics. The characteristics are namely, the probability of reaching the root node & sub attack nodes, the mean time to reach the root node and the mean time spent in the sub attack nodes before reaching the root node. Additionally, the method quantifies the effectiveness of specific defenses in reducing the attack risk considering the efficiency of individual defenses.

-In this paper, the theoretical optimum of a battery utilization in a balancing group for cost reduction and energy imbalance reduction is investigated. To estimate the theoretical optimum, an offline optimization of the battery schedule is performed over a horizon of one week. Different strategies are applied, prioritizing either the cost reduction or the energy imbalance reduction objective to investigate the potential to fulfill both objectives and to determine the trade-off between them. The investigation of the scaling effects of increasing the number of batteries highlights the advantages and disadvantages of the applied strategies. Finally, it is shown that batteries can be utilized to fullfill both objectives to a certain degree, but the potential is limited due to diminishing marginal benefit when increasing the number of batteries.