Probabilistic load flow analysis methods for the integration of PV generation

Uncertainty in distribution grid planning is rising caused by intermittent distributed generation and changing customer behaviour. To adequately consider uncertainty in grid calculations, probabilistic load flow analysis can be applied. In this paper, a case study is presented, which applies probabilistic load flow methods on a low-voltage distribution grid in a rural residential area with a high density of photovoltaic customers. Time series of load data based on smart meter measurements are used to determine probability distributions of grid power infeed and consumption, which are subject to correlation modelling under the use of rank correlation coefficients and a Gaussian copula. The loading of the distribution transformer, respectively its apparent power distribution over one year, is selected as output variable under investigation. The results are compared to transformer loading measurement data as well as a time series calculation based on standardised load profiles. The proposed approach provides promising results and appears to be suitable for application in grid planning, encouraging further research on the topic.