High power quality maximum power point tracking-based islanding detection method for grid-connected photovoltaic systems

Islanding is a condition when distributed generators (DGs) are disconnected electrically from the upstream network. This unwanted situation should be detected effectively to ensure the safety of the maintenance staffs and power quality (PQ) requirements. This paper presents a new high PQ maximum power point tracking (MPPT)-based methodology for detecting the islanding operating mode of grid-connected photovoltaic systems (GCPVSs). In the recommended two-level scheme, a disturbance is injected into the MPPT algorithm under suspicious conditions, recognized by a passive criterion. This disturbance declines the DG active power output remarkably, drifting the output voltage beyond the minimum standard set in islanding state while its impact is negligible at the network presence. The effectiveness of the proposed technique has been evaluated through several hardware-in-the-loop simulations for a case study system, containing two power plant GCPVSs equipped with a pair of multi-functional relays. The results highlight precise islanding classification within 137 ms with the small non-detection zone. Moreover, the results of PQ analyses indicate acceptable total demand distortion and harmonic spectra of the output current in compliance with the existing standards under various DG power penetrations. Since the presented scheme diminishes the active power output in case of suspicious islanding events, its influence on GCPVS efficiency has been studied as well. The outputs underline that the efficiency drops by 0.52% whilst the disturbance is stimulated every minute of the time. It is finally concluded that the proposed technique provides a reliable islanding classification as well as insignificant degradation of PQ and efficiency.

Automated summary

The paper introduces a new methodology based on high PQ maximum power point tracking for detecting the islanding operating mode of grid-connected photovoltaic systems. By injecting disturbance into the MPPT algorithm and applying passive criteria under suspicious conditions, precise islanding classification within 137 ms is achieved.