| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.02 MB | Adobe PDF |
Authors
Abstract(s)
O constante crescimento dos produtores em regime especial aliado à descentralização dos pontos injetores na rede, tem permitido uma redução da importação de energia mas também tem acarretado maiores problemas para a gestão da rede. Estes problemas estão relacionados com o facto da produção estar dependente das condições climatéricas, como é o caso dos produtores eólicos, hídricos e solares. A previsão da energia produzida em função da previsão das condições climatéricas tem sido alvo de atenção por parte da comunidade empresarial do setor, pelo facto de existir modelos razoáveis para a previsão das condições climatéricas a curto prazo, e até a longo prazo.
Este trabalho trata, em concreto, do problema da previsão de produção em centrais mini-hídricas, apresentando duas propostas para essa previsão. Em ambas as propostas efetua-se inicialmente a previsão do caudal que chega à central, sendo esta depois convertida em potência que é injetada na rede. Para a previsão do caudal utilizaram-se dois métodos estatísticos: o método Holt-Winters e os modelos ARMAX.
Os dois modelos de previsão propostos consideram um horizonte temporal de uma semana, com discretização horária, para uma central no norte de Portugal, designadamente a central de Penide.
O trabalho também contempla um pequeno estudo da bibliografia existente tanto para a previsão da produção como de afluências de centrais hidroelétricas. Aborda, ainda, conceitos relacionados com as mini-hídricas e apresenta uma caraterização do parque de centrais mini-hídricas em Portugal.
The continued growth of small electric producers allied to the power injection in various points of the electric grid allowed a reduction of energy imports, but also poses major problems for the grid management. These problems are related to the fact that the production is dependent on climacteric conditions, as example of wind, hydro and solar producers. The energy forecasting according to the weather forecasting has drawn the attention of this business sector, considering the fact that exists reasonable models for short-term, and even long-term, weather forecasting. This work deals, in particular, with the problem of forecasting electric power of small-hydro power plants and presents two solution proposals for this prediction. Both proposals initially predict the streamflow that arrives to the power plant, which is then converted into the power injected in the grid. For the streamflow forecasting two statistical methods were used: the Holt-Winters and the ARMAX models. The two forecasting models proposed consider a forecast time horizon of one week, with hourly discretization, applied to a power plant located in the north of Portugal, named Penide. This work also presents a small study of the existing literature, in matters of power and streamflow forecasts of hydroelectric power plants. It also addresses some concepts related to small-hydro power plants and presents a characterization of the existing small-hydro power plants in Portugal.
The continued growth of small electric producers allied to the power injection in various points of the electric grid allowed a reduction of energy imports, but also poses major problems for the grid management. These problems are related to the fact that the production is dependent on climacteric conditions, as example of wind, hydro and solar producers. The energy forecasting according to the weather forecasting has drawn the attention of this business sector, considering the fact that exists reasonable models for short-term, and even long-term, weather forecasting. This work deals, in particular, with the problem of forecasting electric power of small-hydro power plants and presents two solution proposals for this prediction. Both proposals initially predict the streamflow that arrives to the power plant, which is then converted into the power injected in the grid. For the streamflow forecasting two statistical methods were used: the Holt-Winters and the ARMAX models. The two forecasting models proposed consider a forecast time horizon of one week, with hourly discretization, applied to a power plant located in the north of Portugal, named Penide. This work also presents a small study of the existing literature, in matters of power and streamflow forecasts of hydroelectric power plants. It also addresses some concepts related to small-hydro power plants and presents a characterization of the existing small-hydro power plants in Portugal.
Description
Keywords
Previsão da energia Previsão de produção Centrais mini-hídricas Previsão do caudal Modelos de previsão Energy forecasting Electric power forecasting Small-hydro power plants Streamflow forecasting Forecasting models