Confiabilidade de sistemas de distribuição: calibração de dados e mecanismos para avaliação de desempenho

Abstract

Nowadays, distribution networks operate under a new re-regulated market structure. In this new context, utilities should reduce their costs, which generally occur through measures that may significantly decrease the system reliability. Aware of this problem, the regulatory agencies have created targets for the reliability indices. If any of these standards are violated, penalties are applied to utilities. In order to reduce the amount of penalties, the system must become less susceptible to failures. Thus, reliability evaluations of distribution systems represent important studies nowadays, since they can reveal the weak parts of the systems, or, in other words, those areas that should be reinforced. One of the most relevant techniques used for this proposal is the sequential Monte Carlo simulation (MCS). This method can represent chronological transitions among all possible system operational states and, consequently, provide density probability functions of the performance indices. The methodology to analyze the distribution system behavior should use an amount of reliable database, which represents the reliability aspects of the system components. Therefore, data calibration procedures have to be implemented with MCS technique in order to ensure predicted results with higher degree of credibility. The systems can be better evaluated for their operational condition if a mechanism known as “performance based rates”, PBR, is used. The PBR mechanism rewards the utilities that provide good reliability, and penalizes the ones with poor reliability behavior. PBR contracts introduce attractive investment scenarios to the utilities by offering benefits for those robust systems. The objectives of this research is to present an efficient sequential MCS, propose a new methodology for data calibration based on the probability distribution functions of the indices, and finally discuss two different PBR mechanisms. All proposed methodologies will be applied to a test system, know as IEEE – RBTS, and to a real system.