Projeções de disponibilidade de usinas hidroelétricas através de simulação de Monte Carlo

Abstract

Brazil currently has an electrical matrix made up of approximately 83% of renewable energy and of this source, about 63% is composed of Hydroelectric Power Plants (HPPs). Thus, Brazilian plants have several regulatory issues regarding their contracts, operation and maintenance. A highlight is related to the Availability Factor (AFA), which is responsible for evaluating the availability performance of the plants and their generating units (GUs). The AFA represents an availability target to be met by the generation agents, and a poor performance of the indicator can generate financial penalties. The AFA is mainly influenced by forced stops that occur in generating units. Such stops have a random behavior and directly influence the operation and availability of the hydroelectric plant. Thus, this work presents a Monte Carlo Simulation (MCS) to obtain availability projections of Hydroelectric Power Plants. The main objective of the simulation is to generate forced availability scenarios, helping the agenda in the risk analysis and decision making in relation to the HPP. The proposed methodology consists of two steps. Firstly, the maintenance calendar is optimized, defining the scheduled shutdown days for each plant. Then, the forced shutdown scenarios drawn during the SMC and the scheduled shutdowns are inserted into a model for optimizing the plant’s operation and maintenance. Therefore, through the MCS, scenarios of availability of the HPP will be designed and from that, the risk analysis for the case study is carried out. The history used for the draws is the official National System Operator database, which has data on forced and scheduled events. The proposed modeling was applied using real data from UHE Santo Antônio, which is one of the largest hydroelectric plants in Brazil.