Security Assessment and Protection of Cyber-Physical Energy Systems

The quest of a decarbonized modern society has increased the integration of small Distributed Generation (DG) units that are based on renewable sources of energy in the electric power system, and has entailed a decentralization of power generation. In parallel with a decommissioning of central fossil-fueled power plants, this development has gradually decentralized active and reactive power flow management from the transmission to the distribution network.

Moreover, the increased focus on investment and utilization of the Information and Communication Technology (ICT) infrastructure has caused a transition of the power system into a Cyber-Physical Energy System (CPES). From a distribution network perspective limited attention has been paid to assess the security and protect the CPES operation, therefore new methods are needed to 1) enable the monitoring of Low Voltage (LV) network operational conditions using the existing ICT infrastructure, and 2) protect the DG controls against cyber-physical interdependencies.

The work presented in this thesis focuses on distribution network security assessment and protection in a CPES perspective considering 1) the entailed information security challenges in terms of confidentiality, integrity, and availability, and 2) the limited performance of the existing ICT infrastructure deployed in distribution networks in terms of metering, communicating, and processing equipment.

Acknowledging these challenges and limitations, this thesis proposes two LV network monitoring approaches, a demonstration of a bottom-up approach to DG unit cyber error detection, and an investigation of the impact from cyber-physical conditions on the coordination of DG controls in a LV network. From the activation of the General Data Protection Regulation (GDPR) by the European Commission, the information about consumption from residential load points can be restricted for LV network monitoring.

Therefore, only shared network quantities can be used estimate network conditions. The first LV network monitoring method proposed in this work addresses this challenge. While acknowledging the risk of asynchronous measurement acquisition from smart meters, it estimates the interval of voltage magnitude conditions of an entire radial LV feeder through processing the measurements from network nodes one by one.

If information confidentiality concerns are less stringent, individual household consumption can be applied in monitoring of the LV network. However, existing Distribution System State Estimation (DSSE) methods ignore the time requirements of three phase state estimation for LV network application, which combined with the volatile LV network operation scenarios from demand side management and operation of DG units, can diminish DSSE results.

Therefore, the second proposed LV network monitoring approach considers these challenges, and applies DSSE through a bi-level processing platform. This platform is developed to utilize both periodic and event-driven measurements and can return network conditions during and between DSSE execution. With the decentralization of generation and control from an increased integration of DG units, it is increasingly important to protect the DG controls against hazardous commands.

Their protection is however challenged since both cyber and physical system operational conditions affect the reliability of DG controls. Such challenge is considered in this work through demonstration and investigation of DG control protection against cyber system integrity disturbances, and physical system perturbations.

In particular, a physical system analysis based approach for cyber error detection, is demonstrated through establishment and implementation of state estimation and bad data detection techniques in a distributed processor. This represents a bottom-up approach to DG control protection against unintended and intended information integrity disturbances from ICT infrastructure deficits and cyber-attacks, respectively.

In the end, a cyber-physical simulation platform is established to investigate and coordinate DG control strategies in cyber-physical conditions.