Verification of a Probabilistic Model for A Distribution System with Integration of Dispersed Generation

In order to assess the present and predict the future distribution system performance using a probabilistic model, verification of the model is crucial. This paper illustrates the error caused by using traditional Monte Carlo (MC) based probabilistic load flow (PLF) when involving tap-changing transformers. This is due to the time-dependent feature of tap position, which is not taken into account in the traditional MC method. A modified MC method using block sampling is proposed to take into account the time-dependent feature of the tap position. Block sampling estimates the tap position in the load flow calculation at the present hour by using results obtained from the previous hours. Simulation results show big improvement when block sampling is used as compared to the traditional MC method. Finally, 100 simulation runs of the PLF are performed to further ascertain the accuracy of the results obtained from the developed probabilistic model.