Probabilistic Assessment of Transformer Overcurrent in Distribution Systems with Increasing PV Penetration Levels

The installed capacity of distributed photovoltaic (PV) generation in Brazil has experienced a fast growth since the net-metering bill approved in 2012. In only six years, the installed capacity has grown from 34.5 kW to 570,6 MW. If this trend continues, distribution system operators (DSO) must take the undesired effects of distributed generation into account in their planning studies. However, traditional planning algorithms based on deterministic power flow cannot capture the random aspects of PV generation. Hence, this study proposes a Monte Carlo Simulation approach for evaluating transformer overcurrent probability including real data from Brazilian PV panel market that could be used by DSOs. The proposed methodology is applied to a test system including a 72 kVA. 11/0.416 kV transformer with 55 loads with an aggregated load peak at 60 kW. Results shown that the increase in PV penetration level is followed by an increase in transformer overcurrent probability. Furthermore, it was shown that the 80 % PV penetration level is critical for the test system, as the steep of overcurrent probability increases. A sensitivity analysis is also conducted varying the control strategy of the PV system power factor and the transformer low-voltage level.