Currently, there has been a notable surge in sabotage targeting critical infrastructures, particularly energy hub systems. External interventions on these infrastructures have significantly affected various loads, including electrical and thermal loads. Considering the significance of cybersecurity in energy hubs, there are high expectations for investigating the role of topology in their measures. The aim of this study is to investigate the impact of cyber sabotage on energy management within the demand sector, with a specific emphasis on the topology of an energy hub system. According to the structure of the desired energy hub that has been mathematically modeled, the CPLEX solver is used as a powerful optimization tool with the mixed integer linear programming (MILP) model, which serves as a framework for optimizing the operation cost of the energy hub. The findings reveal the occurrence of false data injection (FDI) into the real network loads. This inconsistency has led to disruption of energy management in the system. Utilizing considered scenarios, the first of them as the base scenario, representing the system under normal network conditions. The second scenario is modeled an attack vector aimed at the electric load, specifically utilizing FDI tactics with an understanding of the network topology. In the third scenario, attention turns to modeling a thermal load attack vector, utilizing FDI tactics with an understanding of its topology. In the final scenario, the emphasis is placed on designing an FDI-based attack vector that targets electric loads within the network, without prior knowledge of the energy hub's topology. Although in all scenarios where the attack is executed, the measuring devices during peak load indicate values lower than the actual ones, the intensity of the attack on the system is considerable in attacks conducted with knowledge of the energy hub topology. Taking into account the importance of the peak load intervals, the maximum changes in electric load during peak hours in the second scenario are compared to those in the first scenario, attributing them to interventions, representing approximately a 17.4 % difference. Similarly, assuming knowledge of the network topology during peak consumption, the impact of the attack on the thermal loads results in changes of 17.17 %. Additionally, in the fourth scenario compared to the first, the maximum changes in electric load during peak hours are approximately 2 %. In the meantime, the optimal cost has remained consistent across all scenarios, rendering the attacks undetectable to the operator and ultimately demonstrating the effectiveness of the proposed method.
Read full abstract