Advanced computational methods for the quantitative analysis of cascading failures in power transmission networks accounting for weather conditions

In this thesis work, we have addressed the need for an accurate analysis of the effects of uncertainties in the safety and reliability of electrical power transmission networks subject to cascading failures. In this context, a novel approach combining uncertainties in the operative and environmental conditions, cascading failures and advanced simulation techniques is developed. The computational framework, composed by various numerical models addressing different tasks, is shown able to capture the environmental conditions variability and to reproduce a realistic network response The first part of this work focuses on estimating the probability of some undesired system state. The novelty of the approach lies in the adaptation of tools typically used in the field of structural reliability to the to the one of power networks. In particular we successful apply an advanced Kriging-based Monte Carlo method coupled with a, here proposed, Latin Hypercube-based sampling scheme. The second part of this work focuses on identifying an approach which can be used for the assessment of proper defined reliability indices in transmission power systems. The novelty of the approach is the quantification of the joint consequences of severe weather events, cascading failures and restoration processes. The system failure probability have been then, successfully evaluated with the simulation technique above proposed.