Probabilistic load flow analysis in Distribution Networks with distributed solar generation

Abstract

This paper presents a method of evaluating node voltages within distribution network using probabilistic load flow (PLF) analysis. It builds on the traditional deterministic load flow (DLF) approach which gives a picture of said network at one specific point in time. The probabilistic approach improves this by taking into account network uncertainties such as fluctuations in household consumption and distributed solar generation. Real household metering and rooftop solar output data was taken from throughout Tasmania distribution network to accurately model these network uncertainties as probabilistic variables. Monte-Carlo Simulation (MCS) is performed using this data on a real section of the distribution network as a case study to show applicability of the developed methodology. The model is validated using historical data from this section of the network. Results show a strong potential for application by Distribution Network Service Providers (DNSP).