Intelligent frequency control for the secure operation of modern power system

Abstract

The modern power system is experiencing several changes to meet the objective to become in a zero CO2 emissions industry. The principal changes are the high penetration of renewable energy sources (RESs) into the power system. As the electrical energy produced by the RESs do not satisfy the technical requirements of the power system (AC and nominal frequency), power electronic converters are used as an interface between RESs and the power system. Therefore, these changes are leading to the power system being dominated by power converter-based technologies. The principal concern of the transmission system operators regarding these changes is the power supply uncertainty coming from RESs affecting the power balance. In addition, the significant reduction of the rotational inertia producing that frequency drop faster and reaches deeper values when a disturbance occurs. These two problems negatively affect the secure operation of the power system.

The integration of energy storage systems (ESSs) to counteract low rotational inertia levels and intermittent power supply of RESs is a reality in several power systems worldwide. Consequently, distributed energy resources (DERs) and ESSs as providers of frequency support services to the power system can enhance frequency stability. Furthermore, enabling DERs and ESSs with fast frequency response (FFR) controls helps counteract the rapid frequency deviation. Therefore, developing novel FFR models and control strategies is one of the most prominent research topics. However, if the action of primary frequency control and FFR are insufficient to re-establish the power balance and limit the frequency deviation, under-frequency load shedding (UFLS) is required to arrest the frequency drop. Therefore, the accurate performance of the UFLS scheme is essential to maintain the continuous operation of the power system and avert possible blackouts.

This thesis aims to create novel control strategies to cope with the challenges created by the reduced rotational inertia in modern power systems and provide frequency support. Therefore, the interest of this research work converges into two main topics—the frequency control provision based on FFR control strategies and frequency control during emergency conditions considering the UFLS scheme. The developed methodologies were assessed on realistic power system models considering the expected low rotational inertia scenarios in the coming decades. The real power systems were modelled in DIgSILENT® PowerFactoryTM using publicly available data, including the dynamic model of battery energy storage system (BESS) and variable-speed pumped-storage hydropower plants (PHPPs). In addition, the Python programming language was used to implement several optimisation algorithms.

This thesis provides a statistical assessment of the time series of frequency, kinetic energy and power demand of the Nordic power system. It unveiled a strong correlation between kinetic energy and power demand during high power consumption hours, opening the door to creating sophisticated prediction models only using the power demand forecast. Furthermore, it demonstrated that the ESSs, particularly variable-speed PHPPs and BESS enabled with FFR control, effectively counteract the reduced rotational inertia. It strengthened the theory of ESSs as reliable technology to provide frequency support services to the transmission and distribution network. Lastly, it highlighted the importance of computing the parameters of each under-frequency relay instead of setting all under-frequency relays with the same set of parameters. The proposed optimal UFLS scheme has proved to be an essential control for power systems with low rotational inertia and high power demand levels.