The steady-state frequency can differ from the nominal value in the islanded operation of AC microgrids. This is because dispatchable generators (microturbines) operate on a frequency droop characteristic, and renewable generators (wind and solar) inject power according to their capabilities. Due to variations in load and power injected by renewable generators, the frequency also varies to ensure power balance. This fact should be considered in the microgrid planning stage so that frequency limits are respected, aiming to guarantee the quality of energy delivered to consumers. When load values and renewable generation are known, steady-state analysis (frequency calculation) can be conducted through deterministic power flow. However, in the face of uncertainties, such powers are modeled based on a statistical distribution, and a probabilistic power flow (PPF) must be executed, with Monte Carlo Simulation (MCS) being the most used method for this purpose (despite the high computational effort required). In this context, this work presents a PPF for islanded microgrids aiming at calculating the statistical values associated with steady-state frequency through Unscented Transformation (UT). UT stands out for using a significantly smaller number of samples, reducing computational effort. Results for a 33-bus microgrid indicate a 99% reduction in computational time while maintaining the accuracy of the results.
