On converter control interoperability in multi-terminal HVDC networks
Loku, Fisnik; Monti, Antonello (Thesis advisor); Norrga, Staffan (Thesis advisor)
Aachen : RWTH Aachen University (2022)
Dissertation / PhD Thesis
Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2022
High-voltage DC links are mainly utilized as interconnectors between asynchronous AC transmission systems and for the integration of renewable energy sources. In order to enable their large-scale integration, today’s predominating point-to-point HVDC links are envisioned to develop into multi-terminal HVDC (MTDC) networks. Considering that they comprise of fast and active components, one of the major challenges in the development of MTDC networks are potential converter control interactions, which can lead to system instabilities and possibly a shutdown of the network. To address these potential converter control interoperability issues, several approaches have been proposed that are based on analytical models of the converters and require detailed information on the converter controls involved in a MTDC network. However, considering the European market structure and its regulatory framework, it is expected that MTDC networks can only be established with the involvement of different converter vendors. In these so-called multi-vendor MTDC networks, the disclosure of the required converter control information between the different vendors is not feasible due to proprietary reasons. To investigate converter control interoperability issues in multi-vendor MTDC networks, this dissertation utilizes the Nyquist and the impedance-based stability criterion. After obtaining the required converter impedances via a frequency-sweep impedance measurement method, a modular analytical approach to derive the equivalent DC network impedance is proposed. The proposed approach is suitable for black-box converter models and allows the calculation of various DC network impedances in a short time with sufficient accuracy. Furthermore, it enables an analysis of numerous network topology combinations, operating points and control modes of the converters. Following this, the worst-case scenarios are identified for a network extension case to a meshed four-terminal network configuration. Moreover, a black-box control tuning optimization approach is proposed in this thesis, under the assumption that the structure and tuning of the controllers already installed in a given network are not known, but that an existing (black-box) network model including the converters and their control system is available. The results show that the proposed method increases the stability margin of the DC network at the critical interaction points in the identified worst-case scenarios. Finally, the proposed methods are demonstrated via laboratory-scaled HVDC demonstrator.