PV Sizing Evaluation for Low-Voltage Distribution Power Systems

This paper presents how to determine the proper size of photovoltaics (PVs) in low-voltage distribution systems. The main objective is to demonstrate the use of artificial neural networks (ANN) for examining the diversity of the systems, in order to give an accurate size of the PVs to be installed. This proposed ANN has been trained with a large number of data; it is able to provide an accurate factor for determining a proper size of the PVs. With the ANN recommended size, the PVs can effectively supply the system loads during active daytime, resulting in lower power loss; implying that the system is being operated at a higher performance. The results also show that when the PVs were oversized, reverse power flow was clearly seen, causing significant power loss. And, if the PVs is undersized, an obvious alleviation of the system's burden is impossible, although the overall power loss was reduced; the majority of loads are still being supplied by the transformer.