A Cumulant-Tensor-Based Probabilistic Load Flow Method

Probabilistic load flow analysis is an important part of grid design, optimization, and operation due to the uncertainties in the power network for both generation and demand, increasingly so for newly integrated technologies including wind power and plug-in vehicles. A reliable, fast, and robust mathematical method for such analyses is a key requirement to help support widespread integration of these new generation and load sources. Conventional deterministic Monte Carlo analyses, though simple in implementation, becomes too slow as networks become more complex. In this paper, a new cumulant-tensor based method is used to assess power flows. Probability distribution functions and reliability indices are generated as final outputs. Furthermore, general correlation between input random variables is included in the analysis. An illustrative 2-bus network is presented along 24-bus IEEE system as case studies, showing the capabilities and increased reliability of the method.