Fault detection, fault location and intelligent protection of multiterminal HVDC lines

Abstract

The need to interconnect different power sources, connect unsynchronized systems, have flexibility in power flow and transport energy over long distances, as well as the new concept of smart gridsh as encouraged the return on investment and research in order to consolidate HVDC transmission as a current response to contemporary demands. VSC converters enable the interconnection of multiple HVDC topologies, resulting in a multiterminal network. However, VSC converters are sensitive to short-circuit currents, thus imposing the need for rapid fault detection and location. It is therefore necessary to develop protection and location algorithms specific for VSC-HVDC system requirements, especially in terms of multiterminal networks. The algorithm needs to be selective and extremely fast, isolating the lower portion of the system in a few milliseconds.In this context, this Direct PhD project aims to study and develop a protection algorithm for VSC-HVDC systems. The algorithm will have as a differential the use of intelligent techniques such as artificial neural networks and fuzzy logic. The use of intelligent tools in HVDC protection systems is still a relatively unexplored topic and with few results, and are the most common algorithms based on the traveling wave phenomenon and techniques based on voltage and current derivatives. Nevertheless the large capacity of pattern recognition and analysis of fuzzy variables gives the intelligent algorithms a unique feature, which can be decisive in the ability to meet the speed and selectivity requirements for protection. (AU)