Probabilistic Evaluation of Photovoltaic Hosting Capacity in Unbalanced Distribution Network via Polynomial Chaos Based Kriging Model

Abstract

With the penetration of distributed photovoltaics (PVs) increasing, evaluating PV hosting capacity (PVHC) becomes more and more important for distribution network operations. Traditional evaluation methods ignore the impacts of voltage unbalance in distribution networks and the results are relatively optimistic. Furthermore, the uncertainty of PVs is also required to be considered in the evaluation. Hence, this paper proposes a probabilistic PVHC evaluation method considering unbalanced distribution network (UDN) operations and variable power outputs of PVs. Firstly, the limits of voltage magnitude (VM) and voltage unbalance factor (VUF) are both considered in the proposed method. A probabilistic violation risk (PVR) index is defined to determine the interval of PVHC in UDNs. Then, a polynomial chaos-based Kriging (PCK) method is developed to calculate the probabilistic indices in PVHC evaluation. The PCK method uses a surrogate model to replace the three-phase power flow calculation in UDNs which significantly reduce the computation burdens of probabilistic PVHC evaluation. Finally, an IEEE 123-bus three-phase UDN is used to verify the effectiveness of the proposed method.