Analysis of State Uncertainty for Distribution System State Estimation

The uncertainty and volatility associated with the increasing number of distributed generation in the lower voltage levels has a direct impact on the state variables occurring in the distribution systems. The monitoring of state variables such as nodal voltage magnitudes and line currents is becoming more and more important and has to face the challenge that predominantly not enough accurate real-time measurements are available. The integration of other data sources, such as forecasts or historical values, requires an understanding of how the uncertainty in these data sources affects the quality of distribution system state estimation. In this work, different types of uncertainty are modeled and subsequently analyzed in Monte Carlo simulations for an exemplary medium voltage network to find the statistical properties of the resulting state variables. The frequency distributions as well as the best fitting probability density functions are determined and presented. Thereby, conclusions are derived about how strongly and in which form state variables are influenced and distorted by uncertain input data. With this contribution methods for distribution system state estimation and their parameterization can be improved in the future.