| In this thesis a method for the reliability evaluation of Distribution Networks is developed. The proposed methodology is of an analytic-heuristic nature. It consists in a mixed technique combining the use of Minimal Cut Set Method with a model based in Markov homogeneous processes of discrete State Space and continuous parameter, including the influence of the weather on the failure rates of the components. To compute the time needed for reconfiguration tasks after a failure, the heuristic technique known as Tabu Search has been applied. This technique is used to optimize these tasks, considering multiple objectives, such as maximize the load that is transferred, minimize the required time, and minimize the power losses of the new configuration.;A modification of the tabu search algorithm used to the determination of the reconfiguration tasks is developed to allow to determine the best possible configuration of a known distribution network depending on the state of its switches (opened or closed). A multiobjective optimization is performed considering the expected number of load interruptions, the expected unavailability, the power losses and the distribution of the load between the branches.;The proposed technique has been applied to several examples of Distribution Networks, including real and extended open-air and underground networks.;Besides, it is proposed a purely analytical methodology to obtain qualitative evaluations of electricity supply indices considering different network structures. This approach is very suitable to perform sensitive analyses of the influence of electric and topological factors on network reliability, which can be used to focus reliability improvement investments and to support network planning, comparing different structures under different circumstances. All significant events leading to supply interruption are simulated and the effects of all events are accumulated on primary feeders, on the contrary of more traditional methods where interruptions are accumulated on load points. This strategy provides a more generic network reliability approach, being its main objective the analysis of homogenous distribution structures. The methodology can be used, for example, to compare different distribution strategies, with their advantages and drawbacks related to service continuity. |
