Dimitrovski R (2021)
Publication Language: English
Publication Type: Thesis
Publication year: 2021
To implement the necessary measures for countering the effects of global warming, i.e. the closing of conventional polluting power plants, massive integration of renewable energy sources, integration of e-mobility, etc., to-day’s power systems are undergoing significant changes. One of the biggest challenges in this transformation represents the decentralized nature of the renewable energy generation, which brings up the issue of electricity trans-mission. The High-Voltage Direct Current (HVDC) transmission, together with the Modular Multilevel Converter (MMC) as its most promising tech-nology, seems to be the preferred solution for transmission of bulk power over long distances and integration of wind and solar resources into the power system.
On the other hand, the shutting down of the fossil-fueled power plants can-not happen immediately. Consequently, there will be many occurrences where the HVDC systems will operate near conventional turbine-generator units. Furthermore, to accommodate the high volatility, it is expected that the power system will often operate near its stability limits. For these rea-sons, it is essential to analyze in detail the nature of the interactions be-tween HVDC systems and turbine-generator (TG) units concerning power system stability.
The torsional interactions are a part of the phenomenon known as Subsyn-chronous Resonances, which in turn presents a power system stability issue. Namely, the interactions between TG units and HVDC links in the subsyn-chronous frequency range can create a high mechanical strain on the shaft that could ultimately result in its breakdown.
The purpose of this thesis is to develop and verify a software framework for the analysis of subsynchronous torsional interactions in AC/DC networks called SNAP (state-space network analysis program), and to apply it on se-lected examples. Following a thorough literature review on the topic of sub-synchronous resonances, the eigenvalue analysis method was selected as the main analytical tool. SNAP is implemented in Matlab® and it is based on the component connection paradigm, which ensures modularity and fa-cilitates further expansion. Together with an algorithm for obtaining the network state equations, excellent flexibility and extensibility to any given network configuration and topology are achieved.
Another important part of the thesis presents the proposed mathematical framework for power system component modeling in the dq frame. This mathematical framework addresses the challenge introduced by the multi-plication of two three-phase oscillating signals that contain an arbitrary number of higher harmonics in the dq frame. Positive and negative se-quence single-phase dq transformation is used in order to account for the zero-sequence components resulting from the multiplication of two oscil-lating signals with different harmonic frequencies. By using three different reference frames to describe each harmonic, this framework provides a way to obtain time-invariant models for any number of harmonics.
A comprehensive study system is implemented both in SNAP and in a com-mercial EMT software suited for the analysis of torsional interactions. The comparison of the results reveals excellent agreement thus validating the accuracy of SNAP. Furthermore, the test cases executed during the valida-tion procedure revealed that adverse interactions between HVDC and TG units might occur under certain conditions. The root cause of such inter-actions can only be analyzed using detailed small-signal and multi-physics models like the ones implemented in SNAP.
APA:
Dimitrovski, R. (2021). Modular Framework for Eigenvalue Analysis of Subsynchronous Interactions in AC/DC Networks (Dissertation).
MLA:
Dimitrovski, Robert. Modular Framework for Eigenvalue Analysis of Subsynchronous Interactions in AC/DC Networks. Dissertation, 2021.
BibTeX: Download