FAULT DETECTION AND LOCALIZATION SCHEMES FOR DC MICROGRID NETWORK

Abstract

A microgrid represents a self-contained and localized low-voltage network integrating various distributed energy sources, such as photovoltaic arrays, wind turbines, fuel cells, and energy storage devices. These components work together to supply power to local loads within a specific geographic area. Microgrids can be designed in AC, DC, or hybrid configurations. While conventional AC technology has historically dominated most microgrids, the increasing utilization of DC distributed energy sources generating DC voltages has spurred interest in DC microgrids as a viable alternative. Despite their advantages, implementing protective schemes and devices in DC distribution systems presents specific challenges. In DC distribution systems, short circuit faults pose particular difficulties. During such faults, the charged DC-link capacitor discharges rapidly, resulting in high-magnitude fault currents with steep rates of change in current ('di/dt' values). As a result, much faster fault detection algorithms are required compared to conventional AC grids. Additionally, interrupting a DC circuit during faults proves more challenging due to the absence of natural current zero crossings, which are present in AC systems. To overcome these challenges, modern technology helps. Enhanced data processing makes it possible to use numerical relays for microgrid protection. These relays offer more flexibility and can even include smart solutions right on the chip. To make sure protection is reliable and fast, choosing the right mathematical tools and algorithms is crucial. Digital relays with high data processing speeds allow us to use time-domain data-based methods, which some utilities already use. This thesis proposes several such algorithms for DC microgrid protection, specifically estimating the parameters of DC cables and analyzing DC current signals using Teager energy and superimposed resistance.