Tienda Universidad de Los Andes

This dissertation presents a method for assessing the vulnerability of a composite power system. It is based on the modeling of failures and repairs using stochastic point process theory and a procedure of the sequential Monte Carlo simulation to compute the indices of vulnerability. Stochastic point process modeling allows including constant and time varying rates, a necessity in those scenarios considering aging and diverse maintenance strategies. It also allows representing the repair process performed in the power system as it really is: a queuing system. The sequential Monte Carlo simulation is applied because it can artificially generate all the aspects involved in the operating sequence of a power system and also because it can easily manage non-stationary probabilistic models. The indices of vulnerability are the probability of occurrence of a high-order loss of component scenario, its frequency and its duration. A high-order loss of component scenario is that one higher than n-2. Examples using the IEEE One Area RTS show how the presence of aging and others factors that produce increasing component failure rates dramatically increase the risk of occurrence of high-order loss of component scenarios. On the other hand, the improvement in aspects such as preventive maintenance and repair performance reduces this risk. Although the main focus of this method is composite power systems, its development produced other outcomes, such as procedures for assessment of power distribution systems, protective relaying schemes and power substations. Carlos J. Zapata