This paper presents a computer model for the probabilistic representation of wind farms generation for reliability studies, which can provide an annual estimation of energy production and calculate several performance indexes. The model combines the stochastic characteristics of wind speed with the operational information of the turbines, such as the failure and repair rates, representing the wind farm by a Markov process. The simulations are made with real time series of wind speed of several Brazilian regions and actual turbine models. The influence of some wind farm and installation site characteristics on the results are evaluated, such as the wind speed statistical clustering technique, the number and type of the turbines, and the failure and repair rates. The results obtained reproduce successfully the behavior of the components considered in the model.

RESUMOEste trabalho desenvolve um modelo computacional de representação probabilística da geração de usinas eólicas para estudos de confi abilidade. O modelo pode fornecer a estimativa anual da energia produzida e calcular indicadores de desempenho, que podem ser usados na análise de viabilidade de implantação das usinas. O modelo combina as caracterís-ticas aleatórias da velocidade do vento às informações operativas das turbinas, tais como as taxas de falha e de reparo, representando o comportamento da usina eólica por um processo de Markov. Foram feitas simulações com séries reais de velocidade de vento de regiões do Brasil e os resultados reproduziram com sucesso o comportamento de todos os componentes presentes no modelo. PALAVRAS-CHAVE ABSTRACTThis paper develops a computer model for the probabilistic representation of wind farms generation for reliability studies. The model can provide an annual estimation of energy production and calculate performance indices, which can be used on wind farms implantation feasibility analysis. The model combines the stochastic characteristics of wind velocity with the operational information of the turbines, such as the failure and repair rates, representing the wind farm by a Markov process. The simulations were made with real series of wind velocity of several Brazilian regions and the results reproduced successfully the behavior of the components considered in the model.

No abstract

No abstract