A robust islanding detection method with zero-non-detection zone for distribution systems with DG

Alexandre Serrano-Fontova, Juan A. Martínez, Pau Casals-Torrens, Ricard Bosch

Research output: Contribution to journalArticlepeer-review


This paper proposes a strategy for detecting unintentional islanding operations (IOs) in distribution networks (DNs) with distributed generation (DG), which eliminating the non-detection zone (NDZ). This hybrid method achieves a zero-NDZ by taking advantage of both passive and active methodologies for an inverter-based DG scenario. The passive-based part of the proposed method considers settings with low thresholds and is activated whenever they are surpassed. The following step uses a three-phase static RC load. This load is connected to intentionally force the frequency and its derivative to exceed the established thresholds. Thus, the events with zero power imbalance can be identified. Unlike other existing methods, this technique does not degrade the power quality (PQ) and does not require DG output power curtailment. The evaluation of the proposed strategy has been carried out through an extensive set of scenarios considering both islanding and non-islanding events. The islanding detection capabilities of the proposed method have been explored considering a custom-made DN test system and the test system recommended by the IEEE 929-2000 standard. The proposed method has a simple implementation, requires a low level of computational complexity, provides a high degree of reliability, and assures fast islanding detection.
Original languageEnglish
Article number107247
Pages (from-to)1-16
Number of pages16
JournalInternational Journal of Electrical Power & Energy Systems
Early online date11 Jun 2021
Publication statusPublished - 1 Dec 2021


  • distribution networks
  • distributed generation
  • islanding detection
  • smart grids


Dive into the research topics of 'A robust islanding detection method with zero-non-detection zone for distribution systems with DG'. Together they form a unique fingerprint.

Cite this