A novel branch-based power flow algorithm for islanded AC microgrids

Abstract

Interest in microgrids is reaching the viral level, fed by the growing research attention and industry investment they are attracting. With its ingrained islanding capability, a microgrid can operate in isolation from the main grid while still keeping its load supplied for enhanced reliability and supply security. These microgrids represent the building block for future smart distribution systems. For steady-state studies of this type of isolated operating mode, an accurate, fast, scalable, efficient power flow analysis tool is crucial. Conventional power flow analysis face challenges when applied to isolated microgrids such as singularity results from system radiality, high R/X ratio, relatively small rated distributed generation (DG) units and thus absence of conventional slack bus, extensive computational burden, convergence issues and extensive solution time. Although branch-based power flow techniques are more preferred for distribution systems to overcome some of the aforementioned challenges, these techniques are incompatible with isolated microgrids as they necessitate the availability of a slack bus. This paper presents a novel branch-based methodology based on the use of power sweeps to solve the steady-state power flow problem. The proposed forward-return-forward-backward sweep (FR-FBS) includes consideration of inherent isolated microgrid characteristics, such as its variable frequency and the absence of a slack bus. The proposed approach also incorporates practical operating modes for DG units and a variety of load characteristics. In addition, the inversion-free nature and consequent speed of the algorithm avoid excessive computational time, thus making it suitable for large distribution networks. The proposed algorithm has been applied to test systems and the obtained results were compared to those obtained with time-domain simulations as well as newton trusted region load flow method. In addition, the effect of load changing on micrgrid’s voltage profile and power sharing among different DGs were studied. The results show the accuracy and effectiveness of the proposed algorithm. This tool therefore offers support for online energy management, self-healing applications, and emerging techniques for clustering power systems into adaptive self-adequate microgrids.