A Multiobjective Voltage Unbalance Factor for PV Hosting Capacity with Probabilistic ZIP Load Models

This paper proposes a multiobjective voltage unbalance factor to estimate single-phase photovoltaic (PV) hosting capacity in 3-phase residential networks. This factor is derived from a combination of various components, including zero and negative-sequence voltage unbalance factors associated with the inherent variability and uncertainty of both loads and PV generation. A time-varying probabilistic ZIP load model is also incorporated in the proposed factor to characterize the uncertainty of voltage-dependent loads within each hour of a day over four seasons. The location, size and phase allocation of PV units are randomly generated. Simulation results obtained on a modified IEEE 4-bus test feeder and the IEEE European low voltage test feeder show that the developed approach can provide a better response to the hourly-varying and uncertain behaviors of loads and generation.

I. Introduction

The popularity of grid-connected solar PV systems has burgeoned exponentially at an annual rate of 18% across Australia in [1]. Particularly, the power grid hosted over 1.6 million rooftop PV systems at the residential level until 2016. Such a tendency is expected to provide a promising remedy to addressing the future energy security. Nevertheless, high PV penetration in single-phase connection would increase voltage unbalance in 3-phase circuits in [2]. The inherent intermittency of solar power along with time-varying and uncertain loads might also have enormous impacts on voltage unbalance, including the zero and negative-sequence components. The detrimental effects of voltage unbalance in 3-phase networks entail the deterioration of the performance and lifetime of equipment (e.g., transformers, induction machines, power converters and power conductors) in [3]–[4]. In addition, random single-phase PV injections in 3-phase 4-wire circuits might lead to further impacts such as neutral line voltage drops, loadability and network losses. Consequently, it is enormously essential to probabilistically quantify the voltage unbalance for single-phase PV hosting capacity. This term is defined as maximum PV capacity that a system can accommodate with no voltage unbalance violations in [5].