Optimisation of battery energy storage systems for enhancing frequency stability of modern power systems under uncertainty

The frequency stability of a power system can only be maintained by keeping the system's generation and loads consumption balanced under all conditions. The increased replacement of Synchronous Generators (SGs) by inverter-based generation sources (IBGSs) such as wind turbines and photovoltaic systems led to frequency stability becoming one of the major issues for system operators. Energy Storage Systems (ESSs) have recently been recognised as a viable technology for various power system stability applications, including frequency stability, due to their fast and dynamic response capability. One of the main advantages of the Battery Energy Storage System (BESS) over the other types of ESSs is that it offers a high level of flexibility regarding the utilised size and placement location. In this thesis, different aspects of BESS participation in the frequency stability of power systems have been studied, including BESS integration structure, BESS sizing/placement, and community-level BESS engagement in frequency stability service. In power systems, different factors can affect the accuracy of the frequency stability assessment, such as ignoring the system uncertainty. Thus, this research initially studied the impact of the system uncertainty on assessing frequency stability under the probabilistic approach. The results showed that considering the system uncertainty provided a more accurate frequency stability assessment when compared to the deterministic analysis (worst-case scenario analysis). The second issue studied in this thesis is related to the integration structure of BESS in the system. In the literature, the difference in the frequency stability support provided by the BESS under different integration approaches has not been clarified, especially for large-scale interconnected power systems cases. In this thesis, the performance of BESS in frequency stability service under the centralised and distributed approaches has been investigated and compared. The obtained results showed that the distributed approach of the BESS provided better frequency stability performance compared to the centralised one. Defining the accurate size and location of BESS in the system is critical, which substantially impacts the grid stability and economy by preventing issues such as over/under-sizing or choosing inappropriate BESS placement. In the literature, most BESS sizing and placement studies rely on optimisation techniques. Also, these studies rarely consider system uncertainty in their modelling. Therefore, novel methodologies have been proposed in this thesis to obtain the optimal BESS size and location while considering a variety of system uncertainty. The results showed that the proposed methods accurately defined the optimal BESS size and placement based on the amount of the system loads and verified that on two different power system models. In recent years, there has been much attention to using neighbourhood BESS (N-BESS) in power systems applications. Since the N-BESS is a grid-connected BESS, it offers a new regime of community-scale batteries. Compared to the other types of community-scale BESS, such as the residential BESS and electric vehicles, the N-BESS comes with a much larger capacity (up to 5 MW). In the literature, most N-BESS studies have focused on services such as improving power quality, increasing self-consumption, peak shaving, and demand management. In this thesis, the role of N-BESS in frequency stability is investigated and compared with large-scale BESS (L-BESS) performance. Also, smart engagement strategies of N-BESS in the frequency stability with and without collaboration with L-BESS have been proposed. The obtained results showed that the proposed strategies enabled N-BESS to participate in frequency stability while maintaining sufficient state-of-charge levels to support the customers' systems in the community.