Search

Power protection systems play a critical role in ensuring the safe and reliable operation of modern power grids. With the increasing complexity of grid topologies and the integration of Distributed Energy Resources (DERs), traditional protection schemes face challenges in maintaining effective coordination among relay systems. This paper presents a new adaptive Overcurrent Relay (OCR) protection and investigates the resilience of different traditional and new adaptive OCR protection schemes against various cyber-physical threats in a smart grid environment. Key contributions include the implementation of a novel adaptive protection scheme designed to dynamically adjust OCR settings based on real-time grid conditions, improving flexibility and responsiveness in managing grid variations induced by DER integration and operational changes. The proposed scheme is validated and implemented using the Multifunction Protection Relay SIEMENS 7SJ62, following to the programming standards of IEC 61131–3, ensuring the reliability and practical applicability of the adaptive OCR scheme in real-world scenarios. Additionally, a dedicated testbed is developed to simulate real-time cyber-physical interactions, incorporating hardware components such as the SIEMENS 7SJ62 and OMICRON-256 test device, along with high-performance computers and communication networks to facilitate comprehensive evaluations of adaptive OCR schemes under varying operational conditions. The study also examines the implications of real-time cyber-attacks on power system operations and the effectiveness of adaptive OCR protection schemes, addressing cybersecurity concerns and proposing mitigation strategies to the system resilience.

Robots are important examples of cyberphysical systems. Typically, robots are battery powered, which are a potential target for cyber-physical attacks to drain batteries and reduce their lifespan. When the battery is drained, the robot is not available and results in denial-of-service. Hence, robotic security and operation duration are fundamental requirements. The main objective of this paper is to provide an impact-based quantitative security risk assessment of three application level attacks targeting a well-known mobile robot platform that is called the PeopleBot™. The novelty of our work is that we successfully drained a fully-charged robot battery using application level attacks that include exhausting the computing resources of the robot. The attacks cause a reduction in the robot availability time. The average availability time from the performed attacks was reduced by 11.78%. We followed the adversarial risk assessment template provided in NIST. Finally, some mitigation strategies for the performed attacks were suggested.

Different technologies are used to manage the integration of smart devices with the conventional power grid. This new integration allows more control and monitoring capabilities for stakeholders and customers. However, it also makes the smart grid susceptible to new classes of cyber-physical threats that have to be analysed, evaluated and mitigated. In this paper, we evaluate the risk of manipulating circuit breakers that connect a power generator to the smart grid. Our main contribution is performing risk assessment of the grid by combining the vulnerabilities of its cyber domain and the transient stability analysis of its physical domain. First, we estimate the probability of compromising the energy control centre (ECC) using a Bayesian network. The ECC can be exploited to manipulate circuit breakers. Second, we analyse the impact of manipulating circuit breakers in the IEEE 39-bus test system. Third, the probability of compromise and its impact are combined to quantify risk. Finally, we analyse the effect of integrating photovoltaic (PV) systems on the stability of the smart grid under the same attack scenarios. The results indicate that integrating smart grids with PV systems can improve resilience even if a cyber-attack succeeds.

In an advanced metering infrastructure (AMI), smart meters may communicate in an ad-hoc fashion to perform functionalities like remote metering and demand response. However, the AMI is susceptible to cyber-physical threats caused by malware such as worms. In this work, we derive a probabilistic model for worm propagation in the AMI. The accuracy of the model is validated using three worm propagation scenarios. This model can be used to estimate the time required to infect Nmeters in the AMI based on the worm propagation technique deployed by the attacker. In addition, we use the derived model to investigate the sensitivity of worm propagation to different parameters such as transmission range, worm size, number and location of target meters. We also demonstrate how an attacker can improve the propagation speed of the worm by modifying the worm's target list. The latter is done by comparing the three introduced propagation scenarios. Improving the worm's propagation speed can amplify the possible physical consequence of the attack.