Local Measurements-Based Backup Protection for DC Microgrids Using Sequential Analyzing Technique

• Published in: IEEE systems journal Vol. 14; no. 1; pp. 1159 - 1170
• Main Authors: Meghwani, Anju, Gokaraju, Ramakrishna, Srivastava, Suresh Chandra, Chakrabarti, Saikat
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.03.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Circuit faults; Communication; Converters; Current derivative; current integral; Current measurement; dc microgrid; Derivatives; Distributed generation; Fault currents; Fault detection; Fault tolerance; Feature extraction; Integrals; local measurement; Microgrids; Overcurrent; Phase comparison; Resistance; Sensitivity
• ISSN: 1932-8184; 1937-9234
• DOI: 10.1109/JSYST.2019.2919144

The presence of phasor measurements in ac systems provides a range of protection techniques based on sequence components and phase comparison to discriminate the fault. However, the absence of phasor measurements in dc system reduces the available alternatives for fault detection. In addition, the presence of low fault tolerant converters, large range of fault impedance and varying grid conditions demands sensitive and selective protection schemes. In this regard, several recent works have suggested a communication-based primary protection due to its high sensitivity to faults. However, with a failure in the communication network, the primary protection will also fail to detect a fault. This paper proposes a backup scheme to isolate the faulty section, even in the case of a communication failure. In the literature, overcurrent- and undervoltage-based backup protection schemes are suggested along with unit primary protection. In the presence of low fault tolerant converters and variable fault resistances, the traditional backup schemes may not work well. This paper proposes a new fault detection method for backup schemes, which utilizes only the locally measured current signal, and uses both derivative and integral characteristics of current to ascertain the occurrence of a fault. The proposed method is capable of detecting the fault accurately and within the required time. The performance of the proposed scheme has been assessed on a ± 600 V TN-S grounded dc microgrid under various conditions using hardware-in-loop simulations on real-time digital simulator.