Possibilistic evaluation of photovoltaic hosting capacity on distribution networks under uncertain environment

Abstract

The large-scale access of distributed photovoltaic (PV) in distribution networks (DNs), if not properly evaluated, brings several problems. The uncertainties with PV outputs and load demand are nonnegligible in PV hosting capacity (PVHC) evaluation. However, not all DNs have sufficient historical data or smart meters to support the accurate probability uncertainty model. This paper proposes a α-Cut-based possibilistic PVHC evaluation method that does not rely on probability density functions. First, to characterize exogenous and endogenous uncertainty, the interval model is applied to model the PV outputs, whereas the membership function is used to model the load demand. Second, the definition of system violation risk (SVR) is introduced to quantify the risk caused by uncertainties. Then, the possibilistic PVHC evaluation method considering SVR is proposed. The test on the IEEE 34-bus system illustrates the effectiveness of the proposed method. The PVHC result is 2–10% of penetration higher than over-conservative results obtained by the traditional approach due to the quantification of uncertainties and the introduction of violation risk. Third, the defuzzification of fuzzy results is discussed, followed by a comparison between the function of crisp value and fuzzy number in the actual decision-making of utilities. The sensitivity of simulation parameters and circuit characteristics is analyzed. Finally, the proposed method is proven on the IEEE 123-bus DN to validate its adaptability to a more extensive system.