The lightning performance of transmission lines is a criterion of utmost importance to be evaluated in transmission system reliability. In order to accurately estimate the expected number of shutdowns, it is imperative to account for the stochastic nature of the lightning phenomenon. In this sense, the most widely used technique to estimate these shutdowns is the Monte Carlo method (MCM). Although the MCM is a trustful method, it has a huge drawback in terms of computational burden. Therefore, this paper proposes the use of an alternative statistical technique, the Unscented Transform method (UTM). This method achieves an approximate non-linear mapping by employing a set of points (associated with input statistical data), which are deterministically selected and weighted. The set of these points and weights is then used to estimate the lightning overvoltage across insulator strings, which represents a significant reduction in the number of samples evaluated in the process in relation to the application from MCM. In this paper, the MCM and UTM methods are applied and compared in the lightning overvoltage evaluation, which is carried out with the help of the Alternative Transient Program (ATP). From the studies conducted, it was found that the application of UTM in place of MCM allowed for significant computational gains while maintaining satisfactory levels of accuracy in the estimation of indicators.
Read full abstract